By Jing Qian, Co-Founder and Managing Director, Center for China Analysis

As we enter the final hours before Presidents Trump and Xi meet in Washington during Xi’s September 23–25 state visit, I think we may need to adjust how we watch U.S.–China summitry.

The era of searching for a “grand bargain” may be behind us. The more realistic question is whether both sides can establish a working equilibrium that makes their domestic political goals achievable.

That brings us back to the language agreed in Beijing in May: a “constructive relationship of strategic stability,” with the critical qualifiers, in Washington’s formulation, of “fairness and reciprocity.”

Beijing is now presenting this not simply as a framework for managing the next summit, but as a possible guide for the relationship for “the next three years and beyond.” Its May formulation described constructive strategic stability as cooperation remaining the mainstay, competition staying within proper limits, differences remaining manageable, and peace remaining the baseline.

The key question now is whether that concept can acquire enough operating substance to guide the relationship.

Where can the two countries identify arrangements that each regards as sufficiently fair and reciprocal, and that therefore make strategic stability more durable?

Against that standard, six things are worth watching.

1. Can Trump–Xi goodwill be translated into lasting mechanisms between the two nations?

Personal diplomacy is critical and matters enormously in both political systems. But personal relationships between leaders are not enough to sustain the world’s most consequential bilateral relationship.

Beijing itself is placing extraordinary weight on the leaders. Chinese official messaging now describes head-of-state diplomacy as both the “compass” and “anchor” of U.S.–China relations.

That makes the institutional question even more important: can strategic direction at the top travel downward through two very different political and bureaucratic systems?

The test is whether leader-level direction produces durable mechanisms: regular economic channels, AI and crisis-management systems, constructive areas of cooperation, implementation processes, and working relationships among the officials who actually manage the relationship.

Interestingly, Beijing is now describing precisely this objective in the economic relationship: moving from “crisis response” toward “institutionalized management.” Chinese official commentary has placed the trade and investment mechanism alongside military, AI, law-enforcement, local-government, business, and Track 1.5 channels as part of a denser network of bilateral mechanisms.

The goal should be to build enough standing machinery that every disagreement does not require leader-level rescue.

That matters as both countries approach major domestic political moments: China’s Fifth Plenum on October 26–29 and the U.S. midterm elections on November 3. The Fifth Plenum dates were formally set this week.

Domestic politics is driving diplomacy nowadays. So the question is whether the two governments can build mechanisms resilient enough to keep operating through domestic political cycles.

2. On economics, look for implementation rather than transformation

Trade will remain important.

Investment may become even more interesting.

But again, I would change the standard by which we judge progress.

Neither side is likely to transform the underlying economic relationship this week.

The more practical question is whether they can make it more predictable and more reciprocal.

And “implementation” itself may be contested: the Chinese readout of the latest Bessent-He talks emphasized “implementing the consensus reached in previous economic and trade consultations.” That raises an important question: not simply what does each side want next?, but what does each side believe the other has already committed to—and has it been implemented?

Can existing tariff arrangements actually be stabilized, now that both sides have more tariff instruments at their disposal?

Can disputes over tariffs and critical minerals be managed without repeatedly escalating into broader confrontation?

Can the emerging trade mechanism be institutionalized in a way that moves economic relations away from repeated crisis management and toward more predictable consultation?

And can the two sides eventually construct investment models in which Chinese capital or technology contributes to American production and employment while legitimate questions involving security, data, ownership, and governance are addressed transparently under U.S. law?

This is where fairness and reciprocity have to become more than diplomatic language.

They need institutional expression—and enough shared understanding of what has actually been agreed to make implementation, as well as measurement of implementation, possible.

3. On AI, can dialogue become a firewall—and quickly?

This may be one of the most consequential areas to watch.

The United States and China will continue competing intensely for technological leadership.

But the risks generated by increasingly powerful AI systems do not respect national borders.

The two sides have agreed to formalize an AI-risk dialogue, with another meeting expected in Shenzhen in roughly two months. The United States has also proposed an AI “incident line” through which the two governments could communicate when serious AI-related safety incidents occur. Beijing has confirmed the broader AI dialogue but has not yet publicly endorsed the specific incident-notification proposal.

That is modest but meaningful progress on building a working mechanism.

One key objective moving forward could be to build a defensive firewall for AI to deal with powerful models already in the hands of rogue actors, and to build mechanisms that help prevent accidents, miscalculations, or uncontrolled interactions between increasingly powerful systems from becoming wider crises.

Over time, that could mean mechanisms for incident notification, crisis communication, human control, and the management of shared risks involving cyber systems, critical infrastructure, biological misuse, and other potentially catastrophic applications.

It may be unrealistic to expect Washington and Beijing to agree to slow technological competition when both regard AI leadership as strategically consequential.

But perhaps this is precisely the point: the two countries do not have to resolve their competition over chips, compute, or technological leadership in order to build defensive mechanisms around risks neither side wants to materialize.

The latest discussions themselves suggest that kind of compartmentalization may be possible: risk communication on one track while competition over chips, compute, and market access continues on another.

Strategic competitors can still build a firewall around the most dangerous risks. And indeed, competition may make those defensive mechanisms more necessary—and more urgent.

4. Taiwan remains the hardest test of strategic stability

There will probably be no convergence on the fundamental political positions surrounding Taiwan.

Indeed, Beijing’s warmer pre-summit language has come alongside an equally explicit reiteration of its political red lines, with Taiwan identified as the most consequential issue in the bilateral relationship.

So this is perhaps the clearest test of what “strategic stability” actually means.

Can both sides maintain their positions and deterrence while reducing the risk of miscalculation, unintended escalation, and misreading of each other’s intentions?

In particular, can Washington maintain deterrence around Taiwan while simultaneously assuring Beijing that its one-China policy is not intended to embolden Taiwan to move toward independence?

Strategic stability does not require agreement—it requires enough understanding and predictability that disagreement does not automatically become crisis.

That requires more than political signaling. It requires defensive mechanisms: reliable military-to-military communication, crisis hotlines, advance notification where possible, clear procedures for managing encounters, and channels that remain open even when political relations deteriorate.

These mechanisms do not resolve the underlying dispute. They help prevent a dangerous incident, miscalculation, or misunderstanding from becoming a wider conflict.

5. Can strategic stability also become constructive?

A relationship organized entirely around managing threats will eventually become defined by threats.

Durable stability therefore needs at least some areas where the two countries can show that interaction can produce affirmative outcomes.

Health may be one of them.

Cancer, rare diseases, early detection, selected clinical and regulatory cooperation, responsible uses of AI in medicine, and AI-related biosecurity risks all offer potential areas for bounded cooperation.

A Chinese discovery that helps cure an American patient is not an American defeat.

An American medicine that saves a Chinese life is not a Chinese defeat.

None of this requires strategic trust.

It requires boundaries, safeguards, and predictability—and, again, fairness and reciprocity.

Perhaps that is also how we begin moving, carefully, beyond unreasonable suspicion.

6. Strategic stability also needs human infrastructure

Government mechanisms are only one layer.

The relationship also needs connective tissue between the two societies: scientists, doctors, students, businesses, scholars, universities, and civil society.

These relationships are not merely the “soft” side of diplomacy. They are part of the hard infrastructure of stability.

Interestingly, people-to-people relations have figured unusually prominently in Beijing’s pre-summit messaging—from Xi’s outreach to students at ten American universities to repeated references to youth, local relationships, the Flying Tigers, and his own long history of engagement with American society. Chinese messaging has emphasized that the relationship’s “hope lies in the people, foundation in society, future in youth, and vitality in localities.”

The fewer legitimate connections that remain between two societies, the greater the danger that each understands the other mainly through security institutions, political rhetoric, and worst-case assumptions.

Not every exchange should be preserved regardless of risk.

But strategic competition without human infrastructure becomes increasingly brittle and dangerous.

Resilience requires both government-to-government mechanisms and society-to-society connections.

That is why bridge builders still matter.

In an era of strategic competition, bridge-building can sometimes sound naïve. But the harder the relationship becomes, the more important serious bridge builders become: people and institutions capable of understanding both societies without romanticizing either, explaining disagreement without amplifying it, and preserving channels through which problems can still be solved.

So my test for the summit is not a grand bargain, but a working stabilization that creates room for constructive strategic stability.

That means:

• Fairness and reciprocity where priorities diverge;

• Mechanisms that translate leader-level direction into mutually agreed, measurable implementation;

• Trade and investment arrangements that create greater predictability and reciprocity;

• An AI firewall and joint defensive mechanisms against shared AI risks;

• Guardrails around the hardest strategic disagreements;

• A positive agenda where cooperation can produce mutual benefit, such as cancer and health collaboration;

• And human infrastructure that keeps the relationship grounded in people—a constant and sober reminder of what all of this is ultimately about.

On September 24, Chinese President Xi Jinping will meet U.S. President Donald Trump at the White House, four months after Trump’s state visit to Beijing in May. That meeting produced a set of economic agreements, a new U.S.-China Board of Trade and Board of Investment, an agreement to open a government-to-government dialogue on artificial intelligence, and a commitment by the two leaders to build what both governments called a “constructive relationship of strategic stability.”

Several consequential elements of the relationship remain unsettled ahead of the September summit. The export-control truce reached in Busan on October 30, 2025—which holds both Washington’s Affiliates Rule and China’s October 9 rare-earth export-control package in abeyance—runs for one year, leaving its future uncertain. Preparations for a bilateral AI risk-and-safety dialogue are reportedly underway, although its timing, participants, and agenda remain unsettled. And, according to some analysis, Washington and Beijing construe “constructive strategic stability” quite differently. The administration has also delayed announcing the proposed $14 billion U.S. arms sale for Taiwan, despite receiving congressional pre-approval in January 2026.

We asked a group of China and U.S.-China relations experts what observers should watch to determine what the meeting actually changes: Which decisions, mechanisms, regulatory actions, political signals—or notable absences—would reveal the direction of the relationship after the leaders leave Washington?

—Jennifer Choo

COMMENTS

The Honorable Dr. Kevin Rudd AC

Much ink has been spilt in recent days forecasting the outcomes of this Thursday’s summit between presidents Donald Trump and Xi Jinping. The truth is that these are hard to predict. So perhaps it’s worth reflecting for a moment on the priorities Xi Jinping will have in mind as he wings his way across the Pacific before touching down at Andrews Air Force Base this afternoon.

The number one priority, at least for senior Chinese officials, if not President Xi himself, will be to avoid any personal embarrassment (let alone humiliation) during his brief sojourn in America’s capital. It’s been over a decade since Xi’s last visit to Washington, during the Obama presidency, and that wasn’t exactly easy sailing. On this occasion, there will be premiums on avoiding the slightest protocol glitch, and having all outcomes as orchestrated and as anodyne as possible. For Xi’s domestic audience, being feted appropriately at the White House, without incident, very much represents ‘Mission Accomplished’. It is also critical for the Party’s message to the country and the world at large that China is now being treated as the co-equal of the United States. For Xi, that really matters. In other words, the visit itself is the message, quite separate to whatever of substance might emerge from it.

The second priority will be to secure President Trump’s affirmation of China’s new framing of the relationship as one of ‘Constructive Strategic Stability’ (中美建设性战略稳定关系). Lest anyone underestimate the significance of this objective within the Chinese system, its Embassy in Washington has just published a diagram of what this phrase means.

This nifty little piece of bureaucratic artwork bears striking similarity to the four compass points of the NATO flag. The northern needle represents ‘cooperation’ while, to the left, we see ‘control’ of strategic differences between the two sides (implicitly, but not exclusively, code language for Taiwan). On the eastern point we find competition, but ‘within bounds’; a reference presumably to the non-lethal domains of strategic competition such as their economic contest, tariffs, export controls, artificial intelligence, the race for foreign policy influence around the world, as well as their underlying ideological contest. Then, if these first three compass points can be aligned, we might then proceed to the southern axis which is ‘peace’ or, at least in the words of the illustration, ‘peace within reach’. And just in case the American audience doesn’t grasp the global significance of it all, the Embassy helpfully reminds us that these two giants together represent more than one-third of the global economy and about a quarter of the world’s population, meaning that – beyond their bilateral relationship – ‘stability’ between these giants of the international system matters for the world at large.

The Chinese system loves what it views as the symmetrical or dialectical alignment of underlying truths. On top of that, these compass points remind us of Xi’s enjoinder to the Chinese nation at large that ‘across east, west, north, south and the centre – across government, the military, society and the academy – the Party leads in all’. This was Xi’s take on a time-honoured aphorism first deployed by Mao himself.

Team China’s third priority is more tangible: to keep tariffs as low as possible. This is where real-world economic concerns come to the fore of President Xi’s agenda. The uncomfortable truth for them is that domestically the Chinese economy is performing poorly. Private domestic consumption growth is weak. Private fixed capital investment remains in the recovery ward. Private residential construction is in its sixth year of recession. The only two pistons functioning in the overall Chinese economic engine are state investment and net exports. State-directed investment underpins China’s techno-industrial strategy, drawing both from the budget and from the state banking sector, where Chinese depositors hold the vast proportion of their savings. As a recent OECD report found, Chinese firms received, on average, three to eight times more in government subsidies as firms in the OECD, with that support accounting for almost 60% of their gains in global market share. As for net exports, China’s merchandise trade surplus last year pulled in US$1.2 trillion from the rest of the world. That number grows quarter by quarter. And despite the American tariff wall, China has managed to continue expanding its export markets by piling into Europe and the Global South. This is sending convulsions through what remains of manufacturing in Europe, as well as in South-East Asia – with the exception of Chinese regional investments in third countries designed to circumvent US trade barriers. In aggregate, net exports represent a central contribution to China’s faltering economic growth. For that reason alone, a central Chinese priority for this visit is to sustain the so-called ‘Busan tariff truce’ which the leaders agreed last October at the margins of the APEC Summit in South Korea. This tariff pause comes up for review this November and, on Friday, US Trade Representative Jameson Greer received a brand new suite of tariffing powers by Trump signing the Lindsey O. Graham Sanctioning Russia and Iran Act into law. The worst case outcome for Xi Jinping would be for Donald Trump to reach for the tariff lever once again, especially if the Europeans were then to follow suit.

Xi’s fourth priority relates to critical minerals and rare earths. Through the Chinese lens, the future of this sector is seen as a counterpoint to US export controls on sensitive technology sales to China – principally, advanced semiconductors. Whereas, viewed through America’s lens, China has fallen short of its undertaking at Busan to normalise the trade in critical minerals and rare earths to the US, Japan and other allies. A return to normal export arrangements was part and parcel of the deal struck in Busan whereby America’s punitive tariff regime would be paused for a period of 12 months. This has not occurred. China may therefore engage in overreach, expecting the Americans to liberalise their export controls on semiconductors as a condition precedent to pulling their restrictions off critical minerals and rare earths supply. This could well blow up in Beijing’s face. Critical minerals are a fundamental priority of President Trump, and efforts to prosecute China’s case are more likely to heighten underlying tensions in the US-China relationship than to cause the Americans to back down.

Fifth, both sides have had extensive discussions on the risks of artificial intelligence. These had been broached during their May summit in Beijing. The essential question at stake then was whether the Chinese accept the premise that there is a need for urgent bilateral action to manage the risks associated with frontier models in the hands of rogue states, rogue actors, or for those models to themselves autonomously attack the social and economic infrastructure of both countries. The outstanding question is whether Beijing shares Washington’s urgency on these questions, or whether China believes its own cyber defences are sufficient to withstand any such assault. One further question is whether China accepts that its own open-source and open-weight models, currently being distributed across the world, constitute the same sort of risk as US-generated proprietary models. Minister of State Security Chen Yixin, in a piece titled ‘Comprehensively Strengthen AI Security Safeguards and Promote the Healthy and Orderly Development of AI’, showed signs of embracing the possibility that these real-world risks may in fact exist, and that China just might be as vulnerable as the United States. Despite the fact that Chen went on to attribute all blame to the US for this problem, Beijing may well be starting to appreciate that this is now a problem for the entire world, including China. What guardrails might possibly be agreed between both countries to reduce this risk remains to be seen. One encouraging development is a statement by US Treasury Secretary Scott Bessent on Monday that both sides had agreed to open a bilateral line of consultation and negotiation on this. We still don’t know whether Thursday’s meeting will put more meat on the bone in terms of real-world progress. It is, at least, positive that a bilateral channel appears to have been opened.

The sixth priority is the hoary old chestnut called Taiwan. President Trump made his views on Taiwan known during an interview with Fox’s Bret Baier in May, stating that: ‘when you look at the odds, China is a very, very powerful big country’ whereas Taiwan was a ‘very small island’; that he did not want to ‘travel 9,500 miles to fight a war’ when Taiwan was only ‘59 miles away’ from the Chinese mainland; that Taiwan ‘stole our chip industry’; and that he was pausing a congressionally-approved arms sale to Taiwan valued at some $14 billion. China, of course, would hope to see this arms package continually deferred and ultimately cancelled. After the Washington Summit, each country will host the other for two further face-to-face meetings this year: first during the APEC summit in Shenzhen in November, and then at the G20 summit in Miami in December. It is an open question whether Trump will continue to postpone any final decision on Taiwan’s arms delivery until this sequence of four visits is concluded. By then, he will know whether China has been helpful on critical minerals and rare earths; and, more particularly, whether it has resumed large-scale US agricultural imports ahead of the mid-term elections. These considerations are likely to deeply influence President Trump’s view of China in general and the Taiwan question in particular.

Finally, there are open questions about China’s negotiating position on what positive and productive advances might be possible for the future of the US-China relationship. This is where commentators often point to rebuilding people-to-people exchanges, primarily student exchanges between both countries. Other positive initiatives that have been floated include efforts to ease restrictions on the flow of two-way foreign direct investment to less-sensitive sectors of each other’s economies. A third set of positive initiatives aggregates around future agreements in public health, cancer research, and the joint analysis of clinical trial data that might bring cutting-edge drugs to market much earlier than would otherwise be the case. Of these three, the latter is perhaps the most promising; not only does it go to the heartland of people’s concerns in both countries about the high rate of cancer mortality, but it also opens up new possibilities for deploying AI to analyse clinical trial data (appropriately anonymised and without exchanging genomic material). In other words, a combined initiative between Chinese and US scientists on cancer research and drug development has some practical prospects for success.

Although the precise landing point in each of these seven areas remains unclear, these are the likely parameters within which the Chinese and Americans will be negotiating. It is impossible to predict what outcomes are likely to stabilise the relationship for the long term. The reality is it could go either way. We should remember that, during the first Trump administration, things went swimmingly at first, including the President’s ‘state visit plus’ to Beijing in November 2017. But within two months, the US National Security Strategy announced a new doctrine of ‘strategic competition’ and the White House launched a tariffs jihad against Chinese exports, capped off two full years later with Trump excoriating Beijing over the ‘Kung Flu’ pandemic.

It remains, therefore, in the balance whether this Thursday’s outcomes in Washington genuinely serve to restabilise the relationship for the future. Or whether China, in fact, overplays its hand, resulting in little by way of substantive outcomes for the United States, thereby significantly contributing to a worsening of the relationship once the midterm elections are done and dusted.

The ball, therefore, lies very much in Beijing’s court.

Neil Thomas

Xi Jinping will arrive in Washington for his summit with Donald Trump hoping to prolong the current stability in U.S.-China relations. Since Xi countered last year’s “Liberation Day” tariff hike with disruptive rare-earth export controls, Trump has prioritized good relations with him and put several planned technology restrictions on hold. That has slowed the expansion of the “derisking” agenda of sanctions, export controls, and targeted economic containment that was a pillar of the Biden administration’s China policy. Their Beijing summit in May saw Trump endorse Xi’s framing of bilateral ties as a “constructive relationship of strategic stability.”

Beijing sees the present moment as a golden opportunity to accelerate industrial self-reliance and strengthen its geoeconomic leverage while limiting further U.S. pressure. Of course, the Trump administration has also authorized a record arms package for Taiwan, placed new restrictions on foreign robots, and accused China of “stealing and copying American AI models.” But many Chinese strategists recognize the U.S. president as a relatively benign figure in Washington when it comes to dealing with Beijing. Indeed, word around D.C. is that more hawkish officials have drafted several proposals to hobble China’s high-tech industries, which are ready to go if Trump sours on Xi.

Trump’s centrality to China policymaking makes his personal interactions and negotiations with Xi especially consequential. China’s rare-earth chokehold may encourage restraint, but Trump could swing back to confrontation if he decides Xi is insufficiently respectful or cooperative. That would be Beijing’s worst-case scenario and would likely trigger retaliation, especially given its heightened sensitivity to foreign policy setbacks ahead of a twice-a-decade leadership reshuffle at the 21st Party Congress late next year, when Xi is expected to secure a fourth term.

Xi will want his visit planned down to the minute, with equal billing alongside Trump and no risk of embarrassment. According to former Belarusian diplomat Pavel Slunkin, during Xi’s 2015 visit to Belarus, then-protocol chief Qin Gang requested a 2:00 a.m. inspection of a museum Xi would tour later that day. His team wanted to know exactly which step of a staircase Xi would reach when some music began playing. No wonder Chinese officials are rattled at the relative lack of preparation for this summit. Such meticulous planning sits uneasily with Trump’s improvisation: In May, he surprised his hosts by inviting several American executives into a meeting with Xi.

Beijing therefore is eager to make the summit a “success” in Trump’s eyes without making significant concessions on issues like AI-model distillation, industrial overcapacity, or rare-earth export controls. Expect progress on Trump’s Board of Trade and Board of Investment initiatives, promises of additional Chinese purchases of U.S. goods, more dialogues, and an extension of the economic truce—although an agreed duration would be moot if Trump decides to pivot on China. If all goes well, the two leaders may end the summit by confirming that Trump will attend APEC in Shenzhen and Xi the G20 in Miami later this year. If things do not go well, we will know much sooner.

Lizzi C. Lee

China’s economy on the eve of the summit is in less dire shape than many had expected. Exports are holding up despite trade tensions, largely because of China’s centrality and ascent across global value chains. The global AI investment boom has also been an unexpected tailwind, lifting demand for upstream metals and equipment and helping pull the economy out of deflation.

So China comes to Washington with a strong hand. Xi has good reason to lower tensions with the U.S., but he will not make concessions on Beijing’s manufacturing ambitions, precisely because those strengths explain the country’s current resilience.

Yet that resilience may itself be a deeper source of vulnerability. A model supported by domestic tech and manufacturing strength and external tech demand is unusually vulnerable to geopolitical risks, including export controls and swings in the global AI investment boom.

Global market pressure, however, can only go so far toward forcing China to balance its economy. China’s imbalance is ultimately a domestic political-economy problem, which is why I am not optimistic that the September summit will produce a breakthrough. Beijing has correctly diagnosed the problem. Xi Jinping himself calls for strengthening domestic demand. But diagnosis is not action. Genuine rebalancing would require a much larger redistribution of income and fiscal resources toward households and local governments. Those are difficult choices, with clear political winners and losers.

Xi is all-in on making China a technological superpower on par with the U.S. While he seems serious, at least rhetorically, about boosting consumption and expanding social welfare, institutional inertia still favors industrial upgrading and strategic sectors. The supply-side agenda remains an unequivocal top priority, and meaningful improvement in domestic demand will realistically be slow-moving. There is also a basic fiscal constraint: How do you ask local governments to spend more on households when many are neck-deep in debt?

This does not mean the U.S. lacks leverage over China. Access to the American market and innovation ecosystem is in fact enormously valuable to Chinese firms that need deep pools of capital and high-value customers. But turning that leverage into an effective mechanism for negotiating with China on its economic model is far more complicated. The sectors at the crux of the problem, especially advanced manufacturing and new energy, are also central to China’s growth model and deeply entangled with national security. These are precisely the issues on which a hastily prepared summit is unlikely to produce substantial progress.

What happens beyond September is more consequential: whether China’s need to rebalance can be aligned with America’s need to rebuild industrial competitiveness. Chinese cost-effective tech products, produced through intense domestic involution, could eventually be leveraged against supply-chain localization requirements for Chinese firms to access the U.S. market and reshore jobs and technological know-how. But getting that architecture right requires serious preparation, especially below the presidential level. In that sense, the proposed U.S.-China Board of Investment could eventually be useful, especially if it provides Chinese firms with political shielding to invest in the United States beyond individual election cycles.

So I am not optimistic about a grand deal this month. But signs that both sides are willing to build the machinery for a deeper conversation on building healthier, more balanced economic models would matter. Leadership-level intent still has to translate into bureaucratic incentives on both sides. No single summit can do that in one fell swoop.

Craig Allen

Both the U.S. and Chinese governments are signaling that expectations should be lowered for President Xi Jinping’s September 24 state visit to Washington.

CEOs will pay attention to the Board of Trade, the Board of Investment, announced deals and extension of the bilateral export control “truce.”

We should expect announcements concerning the Board of Trade. For example, there has been discussion of the U.S.$30 billion “reciprocal tariff reduction framework.” This would allow a subset of trade to flow with minimal tariffs between the two countries. The Chinese side is expected to grant the U.S. side duty-free preferences for American agricultural product exports. The U.S. side will likely allow Chinese consumer goods to flow duty-free into the U.S. Outside of the $30 billion lower tariff “valley,” all other trade will be assessed a higher duty “plateau.” The exact composition of the low tariff goods “valley” will significantly impact supply chains.

During the summit, we will likely also learn more about the next steps for the Board of Trade, including its structure, schedule, and operating procedures. CEOs will be interested in private sector engagement with the organization and whether it may prove an effective mechanism for addressing market access problems.

Thus far, there have been very few mentions by either government of the Board of Investment, which was formally announced in May, when President Donald Trump visited Beijing. This mechanism is to consider expedited CFIUS approvals for Chinese investors in the United States. At the summit, CEOs will be looking to see if any investments might be announced or “green lit” for future approval.

As a symbol of bilateral efforts to address the trade imbalance, there may be commercial signing ceremonies for American exporters of aircraft, agricultural, or energy exports. These signing ceremonies may be an integral part of the official meetings, or they may be on the margins of the official events.

CEOs in technology industries will be interested to see if there is an extension of the export control “truce” agreed to at APEC last year. Under the terms of the truce, both sides agreed to postpone implementation of new and expansive export control measures for one year. To date, both governments have myriad complaints about the other. The U.S. is concerned about lack of access to rare earth elements. The Chinese are concerned about lack of access to aircraft parts and recently announced FCC regulations. Both sides blame the other for the impasse.

From a legal perspective, the export control dialogue between the two countries should be about national security issues. In fact, the export control dialogue has often become a discussion over commercial chokepoints and industrial policy. Both sides are seeking maximum leverage and trying to minimize their own vulnerabilities. While the two sides may very well extend the truce for another year, the “supply chain war” will likely continue for the foreseeable future.

CEOs will be watching closely.

Paul Triolo

When Presidents Trump and Xi meet on September 24, the primary deliverable will not be soybeans, Boeing planes, or tariff reduction. It will be an extension of the Busan agreement, probably for one year, because both sides recognize the economic and political fallout that would follow without one. The agreement in Busan revolved around holding in abeyance both the U.S. Affiliates Rule, which would massively expand the reach of the U.S. Entity List, and a new set of export controls on critical minerals, rare earths, magnets, and related technologies that would significantly complicate current and future supply chains for these critical inputs. More important, however, is what each side agrees to restrain—and whether the current technology truce can become more durable.

Four years after the United States pioneered a new level of economic coercion in the name of national security with the October 2022 export-control package, the technology relationship has evolved into a reciprocal system of pressure and restraint. Whatever one’s assessment of the controls’ original objectives, they have not prevented Chinese companies from continuing to develop advanced AI, high-performance computing, and semiconductor capabilities, while Beijing has developed increasingly consequential forms of leverage of its own.

Front and center are the critical mineral, rare earth, and magnet controls Beijing has rolled out over the past two years. Measures covering gallium and germanium, graphite, rare-earth processing and magnet-making technology, and antimony culminated in a further series of actions on October 9, 2025, including Announcement 61 which adds extraterritorial and de minimis controls on rare earths. Announcement 61 requires foreign exporters to obtain a Chinese license for specified foreign-made rare-earth products incorporating at least 0.1 percent by value of designated Chinese-origin inputs, as well as covered items made abroad using specified Chinese rare-earth technology, for which no de minimis threshold applies. Those measures remain in abeyance while the Busan agreement holds. The United States, in turn, has held off on the Affiliates Rule, which would sweep thousands of additional Chinese entities under Entity List restrictions.

That makes actual regulatory decisions, rather than summit language, the first thing to watch. Does Washington continue to hold back the Affiliates Rule or other new technology restrictions, such as those recently implemented by the FCC? Does Beijing continue suspending its October 9 measures and facilitate rare-earth and critical-material licensing? And do the two governments clarify what each understands the Busan agreement to require? A nominal extension that leaves both sides repeatedly claiming that the other has violated a vague understanding would do little to break the current tit-for-tat cycle.

A major unanswered question is whether the two governments can begin to build machinery capable of managing that cycle. There is a real deal to be made that would help with this, as I outlined on my Substack, but it seems unlikely, given political pressure in D.C. to be tough on China. Jing Qian and I have proposed a Board of Technology Stability to bring order and sustained high-level attention to issues that can no longer be left to fragmented bureaucracies unable to gauge the impact of their policies on U.S. and allied companies, global supply chains, and the bilateral relationship. Whether or not that particular mechanism is adopted, some empowered process for handling technology-control disputes before they escalate would be significant.

The summit therefore offers a relatively simple test. Extending Busan would preserve the truce. Concrete mutual restraint in licensing and regulatory implementation would make it more credible. A standing mechanism for managing future disputes would suggest something more durable: movement toward what Daleep Singh has called a “shared doctrine of restraint.” Without those steps, the forces pushing both sides toward further controls, de-risking, and retaliation will keep the technology competition—and the broader relationship—on unstable ground, a major negative for global companies and supply chains.

Helen Toner

The last time the United States and China formally met to discuss AI risk and safety was in Geneva in May 2024, when many of the technology’s security implications were largely hypothetical. By 2026, the hypothetical has become real—but it remains to be seen whether that is enough to produce more productive dialogue between the United States and China.

The Geneva talks in 2024 were mostly notable for having happened at all, during a period of high bilateral tensions. Later that year, then President Joe Biden and Xi Jinping did manage to agree on the importance of human control over nuclear weapons, but otherwise the sense on both sides was that there was little momentum behind any discussions around AI safety and risk management.

This year, things may be different. In March, reporting on U.S. strikes in Iran documented the central role being played by AI in the military’s targeting process. In April, the U.S. company Anthropic announced a new model called Mythos that can carry out sophisticated offensive cyber operations. And in July, a swarm of hundreds of autonomous agents escaped OpenAI’s research infrastructure to attack the AI platform Hugging Face, roaming freely on the open internet for days before being noticed and contained.

No surprise, then, that AI risk management is near the top of the agenda for Trump and Xi. But it remains to be seen whether a new round of U.S.-China discussions on AI safety and security will bear fruit.

One challenge is the sheer breadth of AI-related topics. On the one hand, the fact that AI touches on every industry and domain of society raises the salience of the technology overall, but on the other, it makes it challenging to home in on specific topics of focus. What’s more, there is AI-specific distrust on both sides. The U.S. side is unconvinced that China takes the risks of AI catastrophe seriously, while the Chinese side doubts U.S. sincerity given what it sees as an inadequate regulatory regime in Washington and a history of attempting to slow China’s technological progress.

To determine whether the Trump-Xi summit is fruitful, we can look for two key pieces of groundwork for future talks. Do Trump and Xi select consistent interlocutors to continue discussions at the working level, and can they identify concrete challenges related to AI that both sides want to tackle?

First, selecting the right interlocutors on both sides has been an ongoing difficulty, given the alphabet soup of agencies in each country whose responsibilities touch on AI. If we emerge from this round of engagement with clear counterparts empowered to continue AI discussions at the working level, there will be much better chances of ongoing progress.

Second, a crucial ingredient will be narrowing the discussion from “AI security”—or, in Chinese, anquan (安全), which encompasses both safety and security—to the specific challenges both sides wish to address. There are many options for topics: bad actors’ use of AI for cyberattacks and bioterrorism, risks from openly released AI models, or the industry’s headlong rush towards fully autonomous self-improving AI, among others. Vague statements about addressing risks and promoting benefits, however, would be a sign that meaningful progress is still elusive.

Given the speed at which AI is advancing, the two countries may not have many more chances to get off on the right foot.

Lyle J. Morris

Washington’s long-standing policy toward Taiwan will inevitably test Beijing’s “red line” on Taiwan. This is because U.S. policy, including arms sales and congressional visits to Taiwan, is rooted in the Taiwan Relations Act and reflects enduring, bi-partisan practices that will continue regardless of who occupies the White House.

That policy will by default upset Beijing, which seeks curbs on U.S. arms sales and official engagement with Taipei. Despite having described Taiwan as “the core of China’s core interests” and the “first red line that must not be crossed” in U.S.-China relations, Xi Jinping is not likely to have seen Chinese concerns over U.S. Taiwan policy assuaged when he visits the White House.

The proposed $14 billion U.S. arms package for Taiwan, which Congress pre-approved in January 2026 but which the Trump administration has yet to formally notify to Congress, represents an unusually concrete test of the “constructive strategic stability” framework agreed to by both leaders in May. As of early September, the Trump administration has not formally notified Congress that he has approved the sale. Consequently, neither the package’s final value nor its precise composition has been confirmed by the State Department.

The package reportedly includes counter-drone capabilities, an integrated battle-command system, and medium-range munitions. A bipartisan group of senators urged Trump to formally notify Congress of the package before his May meeting with Xi. Trump did not do so and subsequently acknowledged discussing Taiwan arms sales with Xi “in great detail,” while saying that no commitment had been made and that a decision would follow.

The timing of any notification would nevertheless carry considerable weight. Notification before Xi’s visit to the White House on September 24, which is unlikely, would demonstrate that U.S. security policy toward Taiwan remains insulated from summit diplomacy. Holding the package through the summit could indicate an effort to avoid introducing a major irritant while the two leaders seek broader stability.

Trump should notify Congress shortly after the visit. Doing so would preserve the diplomatic atmosphere while ultimately upholding longstanding U.S. commitments to Taiwan. If Trump were instead to wait longer, that would generate greater uncertainty about Washington’s commitment to Taipei and give rise to suggestions that Trump is currying favor with Xi at the expense of relations with Taiwan.

Make no mistake, Taipei desperately needs the weapons. People’s Liberation Army aircraft and naval vessels continue their relentless coercion around Taiwan. Beijing has expanded its gray zone activities as well, recently deploying Naval and Coast Guard vessels east of the island, characterizing the activity as a “lawful exercise in jurisdiction.” As a result, Taipei’s deterrence posture for both highly kinetic contingencies and lower thresholds is deteriorating. These intrusions are unfolding alongside persistent tensions in the South China Sea and expanding regional security cooperation driven partly by concerns over Chinese military pressure.

The larger question for the summit, therefore, is not whether Washington avoids crossing Beijing’s Taiwan red line. The two governments fundamentally disagree over where that line lies. It is whether they can establish clearer guardrails around that disagreement. “Constructive strategic stability” will acquire meaning only if Washington and Beijing can continue pursuing incompatible Taiwan policies while developing mechanisms that prevent those differences from producing unintended crises.

Jing Qian

When Donald Trump and Xi Jinping meet in Washington on September 24, attention will focus on trade, technology, and Taiwan. But there is a quieter test worth watching: whether the two governments can name a single area where cooperation delivers something Americans can actually see.

In May, the two leaders agreed to “build a constructive relationship of strategic stability.” The White House added an important qualifier: “on the basis of fairness and reciprocity.” The field of medicine is one area where that principle could be demonstrated.

This is not an argument for setting competition aside. The United States and China will keep competing hard in artificial intelligence and biotechnology, and raising legitimate concerns about intellectual property, data security, supply chains, and dual-use technology. Moreover, decades of scientific engagement helped China build research, clinical, and commercial capabilities that now rival—though in some areas complement—American strengths. That’s an argument for approaching cooperation carefully, not for pretending China lacks capacity or for forgoing opportunities to use it selectively where it can strengthen American competitiveness and improve health outcomes for American patients.

Cancer makes the case. The American Cancer Society projects that 626,140 Americans will die of the disease this year—about 1,720 every day. A Chinese discovery that helps cure an American patient is not an American defeat, and an American medicine that saves a Chinese life is not a Chinese defeat. The measure of success should be whether people live longer and healthier lives.

The two countries bring different strengths. The United States leads in basic biomedical science, regulatory standards, capital markets, and commercialization. China has developed extraordinary clinical scale, faster patient recruitment, and a growing role in global drug development. The task is to use those complementary strengths selectively, without compromising either country’s legitimate security interests.

There are practical places to begin. The two sides could work together to speed drug discovery with responsible AI and make clinical evidence more interoperable in cancer and rare diseases, while keeping patient-level and genomic data protected. They could run parallel regulatory work on selected cancer therapies, each regulator retaining its own authority. They could cooperate on early detection, where larger patient populations can help reveal more quickly what works.

And as AI grows more powerful in biology, both countries have an interest in keeping terrorists and criminals from exploiting it to develop bioweapons. That last point matters beyond health. Cooperation today cannot rest on strategic trust that does not exist. It can rest on shared vulnerability to build predictability: narrow commitments, clear rules, notification mechanisms, and evidence that each side does what it said it would. Predictability is not trust. Over time, it can help carry both sides beyond unreasonable suspicion.

The political space exists. In May, Xi named health among the areas where the two countries should expand cooperation. And the calendar is short: China hosts APEC in November, followed by the U.S.-hosted G20 in December, with both governments already committed to supporting the other as host. That gives September a simple task: Turn language into something people can see, with two opportunities soon afterward to demonstrate progress.

Cancer cooperation will not resolve the rivalry. Precisely because the rivalry is durable, both countries need places where working together remains possible.

Health should be one of them.

This special issue of PLA Watch examines a consequential shift in the composition and geography of Chinese military and gray-zone pressure around Taiwan in 2026. We document what appears to be a sharp decline in PLA air activity, with aircraft sorties falling by roughly half year on year. This is juxtaposed with a major uptick in Chinese “official-vessel” activity, such as coast guard and maritime research vessels. When digging into the aggregate maritime data, we find that the rise is driven principally by China Coast Guard activity, in particular operations east of Taiwan’s main island since June 2026. This extends a gray-zone pattern previously concentrated around Kinmen and Matsu, Pratas, and Itu Aba. Finally, we examine Beijing’s effort to routinize these operations through the language and practice of “normalized law-enforcement patrols,” suggesting that China is not simply shifting pressure from the air to the sea, but broadening the geographic scope and legal-administrative character of its coercion around Taiwan. Taken together, these developments point to a changing coercive strategy in which persistent maritime law-enforcement activity is becoming an increasingly important instrument for asserting Chinese jurisdiction and shaping the operational environment around Taiwan below the threshold of overt military conflict.

Summary and Key Judgements

Overall Summary

PLA aircraft sorties around Taiwan in 2026 appear to be operating at roughly half last year’s rate, while naval activity remains generally the same. However, “official-vessel” activity compiled by Taiwan’s Ministry of National Defense has risen sharply. The increase is driven principally by China Coast Guard (CCG) activity, and since June 2026, its geographic center of gravity has expanded to waters east of Taiwan’s main island. The result is not a one-for-one transfer of pressure from air to sea, but a consequential change in the composition, geography, and legal-administrative character of Beijing’s coercion.

Data Scope and Sources:

Summary of Key Judgments

1. Air activity has halved, at both ends of the distribution. Sorties fell 51.1% year on year, from 3,603 to 1,763 over the same 231-day window. Days with no aircraft rose from 12 to 43, while days above 50 sorties fell from eight to zero. The routine baseline and the surge activity visible from Taiwan Ministry of National Defense (MND) reporting both contracted.

2. The character of air activity did not change. The share of sorties crossing the median line or entering Taiwan’s ADIZ was 71.3% in both years. The change is in frequency, not a shift toward “fewer but more aggressive” sorties.

3. Law-enforcement vessel presence more than doubled and became routine. MND official-ship detections rose from 183 to 457 (+149.7%). Days with five or more official ships rose from one to 33. From June onward, the average was about 3.9 official ships per day, compared with 0.9 in January–May.

4. The MND category masks the agency driving the increase. MND groups non-naval PRC government vessels into a single “official vessel” category. Taiwan Coast Guard Administration (CGA) records disaggregate that category into CCG, Maritime Safety Administration (MSA), and fisheries-enforcement vessels; the CCG makes up by far the largest share. This separate evidence supports attributing the maritime law-enforcement increase principally to the CCG.

5. The geographic center of gravity of PRC law-enforcement gray-zone activity has shifted eastward. Since June 2026, CCG formations have operated east of Taiwan’s main island, extending a pattern previously concentrated around Kinmen/Matsu, Pratas (Dongsha), and Itu Aba (Taiping Island). The sequence suggests geographic expansion of an established gray-zone model rather than an isolated episode.

6. Beijing is attempting to normalize an enforcement presence east of Taiwan. On July 4, a CCG formation led by the cutter Xiushan relieved the Daishan formation and Beijing used the phrase “normalized law-enforcement patrols.” This routinization language resembles terminology used in other contested maritime spaces where China seeks to assert jurisdiction through persistent law-enforcement presence.

7. This is a shift in composition and geography, not a transfer of pressure at constant volume. The official-ship increase of 274 detections offsets only 14.9% of the 1,840-sortie air decline. Total observed activity across aircraft, PLAN vessels, and official ships fell 30% year on year. The stronger conclusion is that the mix and operating pattern have changed.

Analysis

Air activity: Volume Halved, Character Unchanged

The year-on-year decline in PLA air activity is large and consistent across the principal measures in Taiwan’s daily reporting. Sorties fell from 3,603 to 1,763 between 1 January and 19 August, a decline of 51.1%. Median-line crossings and ADIZ entries fell by essentially the same proportion, from 2,570 to 1,257. The decline was already emerging in the fourth quarter of 2025: October was the weakest month of that year, and excluding the two-day surge on 29–30 December, the remainder of December averaged 9.6 sorties per day, close to the 2026 baseline.

The most analytically important continuity is the proportion of air activity entering contested space. As Figure 1 indicates, the median-line/ADIZ share was 71.3% in both years. This stability argues against interpreting 2026 as a deliberate trade of sortie volume for greater per-sortie aggressiveness. Beijing is flying less often, but the observed geographic character of those flights is broadly unchanged.

Figure 1: PLA aircraft sorties detected around Taiwan, by month, 2025 vs 2026

Notes: A sortie is one flight, not one aircraft. 2026 data run to 19 August. Over the same 231-day window (1 January – 19 August), sorties fell from 3,603 to 1,763 (−51.1%); the share crossing the median line or entering the ADIZ was 71.3% in both years. The December 2025 figure is carried by two days: 29 and 30 December account for 207 of that month’s 486 sorties, and the remaining 29 days average 9.6 sorties — already at the 2026 level.

Composition Shifts in Maritime Activity

Against the halving of air activity, the maritime picture is more stable in absolute terms but substantially more important in relative terms. As Table 1 illustrates, PLAN vessel detections fell only 4.9%, from 1,706 to 1,622. Official-ship detections, by contrast, rose from 183 to 457. As a share of all reported detections, aircraft fell from 65.6% to 45.9%, while PLAN and official vessels together rose from 34.4% to 54.1%. For the first time in this comparison, most of what Taiwan’s military reports observing around the island is on the water rather than in the air. Figure 2 below compares official ship detection rates from 2025 and 2026.

Table 1: PLA Aircraft Sorties, PLAN vessels, “official” vessels, and overall detections, 2025 vs 2026.

The change in official-vessel activity is not simply a higher count. It is a change in operating pattern. In the 2025 window, only one day recorded five or more official ships; in 2026, 33 days did. The rise began around 28 May and ramped up over roughly ten days. Monthly totals then remained elevated through June, July, and August, turning what had been an episodic presence into something closer to a standing one.

Figure 2: PRC official ship detections, daily average by month, 2025 vs 2026

Notes: “Official ships” is the Ministry’s category for PRC government vessels operated for non-commercial purposes; the daily reports do not identify the operating agency. Daily averages divide monthly detections by days in the month. Over the same 231-day window, detections rose from 183 to 457 (+150%), and the number of days with five or more official ships rose from 1 to 33. The shift began around 28 May 2026, when daily counts moved from 0–1 to 3–4 and then to 5–10 through the first week of June. 2026 data run to 19 August; January–April 2026 values are provisional pending confirmation against the primary compilation.

CCG Pressure Moves East

The additional CGA evidence changes the interpretation of the maritime increase in an important way. Prior PRC law-enforcement gray-zone pressure had been concentrated around Taiwan’s offshore holdings and other contested spaces—most notably Kinmen and Matsu, then Pratas (Dongsha) and Itu Aba (Taiping Island). Since June 2026, that pattern has expanded to the eastern side of Taiwan’s main island. CommonWealth Magazine tracking cited in the CCG analysis found no comparable earlier activity east of Taiwan in the prior MND statistics; the June onset therefore marks a geographic expansion of the existing gray-zone model.

This eastward shift matters because waters east of Taiwan have traditionally been strategically significant as an avenue for resupply, reinforcement, and access to the wider Pacific. A persistent PRC law-enforcement presence there broadens the geography in which Beijing can practice surveillance, jurisdictional assertion, merchant-vessel hailing, and other coercive activities below the threshold of overt military force. It also reduces the analytical usefulness of treating gray-zone pressure around Taiwan primarily as a phenomenon of the Taiwan Strait or the offshore islands.

Legal-administrative Coercion and the Logic of Routinization

Taiwan’s cross-strait maritime regime distinguishes “prohibited waters” (禁止水域) and “restricted waters” (限制水域), corresponding broadly to territorial-sea and contiguous-zone functions but with specially delimited polygons around offshore islands close to the mainland. This framework is operationally important because Chinese law-enforcement incursions can be tracked not only as vessel presence but as challenges to Taiwan’s exercise of jurisdiction in specific maritime zones.

The June–July 2026 sequence suggests Beijing is attempting to establish a new routine rather than conduct a single signaling operation. After a Japan–Philippines summit on 28 May announced EEZ and continental-shelf delimitation talks, China protested and on 1 June dispatched a CCG formation led by Daishan for “law-enforcement” patrols east of Taiwan. From 6–9 June, four large official vessels conducted a special maritime-traffic enforcement operation, entered restricted waters southwest of Cape Eluanbi, and hailed foreign merchant vessels. On 4 July, a formation led by Xiushan relieved Daishan, and the CCG described the activity as “normalized law-enforcement patrols.”

The Japan–Philippines delimitation talks are therefore best understood as the proximate trigger and public legal pretext, not necessarily the underlying strategic driver. The broader pattern—from Kinmen in 2024, to Pratas and Itu Aba in 2025–26, and now east of Taiwan—suggests a sustained effort to widen the geographic scope of PRC law-enforcement jurisdictional claims and make repeated presence appear ordinary.

Analytic Implications

The center of gravity has shifted, but aggregate pressure has declined. The data support an initial conclusion that the observed mix of coercion has tilted from the air to maritime domains.

The CCG is increasingly central to day-to-day coercion. Separate Taiwan Coast Guard Administration (CGA) evidence resolves an important limitation in MND reporting and indicates that coast guard activity is the principal source of the official-vessel increase.

The eastward expansion is strategically more significant than the raw vessel count. Persistent CCG operations east of Taiwan extend jurisdictional gray-zone tactics into a maritime space with direct relevance to Taiwan’s links to the Pacific and to key sea lines of communication (SLOCs).

Routinization may be the objective. The use of “normalized law-enforcement patrols,” rapid replenishment of one CCG formations by another, and sustained elevated counts suggest an effort to convert episodic incursions into a new accepted baseline.

The pattern should be monitored as a sequence, not as isolated incidents. Kinmen, Pratas/Itu Aba, and now waters east of Taiwan exhibit a common logic: repeated law-enforcement presence, legal or administrative justification, and normalization of PRC jurisdictional behavior below the threshold of war.

Timelines, Notes and Sources

Timeline: May–July 2026

What MND’s “Official Vessel” Category Conceals

MND’s daily reporting does not identify the operating agency behind “official ships.” The category follows Taiwan’s legal distinction between foreign military vessels and other foreign official vessels and can include CCG ships, Maritime Safety Administration vessels, and fisheries law-enforcement ships. MND data alone therefore cannot support a specific claim that every additional official-vessel detection is a Coast Guard vessel.

Taiwan CGA records provide the missing disaggregation. The CGA data separate CCG, MSA, and fisheries-enforcement vessels and show that CCG vessels account for by far the largest share of the activity. Read together, the MND and CGA datasets sharpen the original finding: the rise in official-vessel detections is not merely a generic increase in Chinese government shipping but is principally an expansion of coast-guard-centered gray-zone activity.

What the Evidence Does Not Show

The datasets remain observational. MND detections measure what Taiwan’s military observed and reported; they do not directly measure PRC capability or intent. The 51.1% fall in detected air sorties is consistent with reduced activity but does not rule out changes in activity outside the reporting series. Likewise, the CGA disaggregation supports identifying the CCG as the principal contributor to the official-vessel increase, but the strategic purpose attributed to that activity remains an analytic judgment rather than a direct measurement.

Sources and Methodological Note

Taiwan Ministry of National Defense daily reports on PLA activity in the waters and airspace around Taiwan, 1 January 2025–19 August 2026; authors’ compilation with complete daily coverage. Taiwan Coast Guard Administration public records, used to disaggregate MND’s “official vessel” category. CommonWealth Magazine reporting and vessel-track analysis on PRC gray-zone activity around Kinmen and Taiwan, including its 4 August 2026 investigation of CCG operations around Taiwan. The quantitative comparisons preserve the original 231-day year-on-year window.

The devastating floods along the China-Nepal border in late August highlight the growing political importance Beijing places on disaster response. After a glacier collapse triggered catastrophic flooding and landslides across Nepal and Tibet, President Xi Jinping ordered an “all-out rescue effort.” This is significant as during his nearly 14 years in power, Xi has rarely appeared at the frontline of a disaster, in stark contrast to his predecessors Jiang Zemin and Hu Jintao.

On August 28, Xi chaired a Politburo meeting focused on disaster response, while Premier Li Qiang travelled to Gyirong County in Tibet to oversee relief efforts. Both called for stronger monitoring and early-warning systems, as well as measures to prevent secondary disasters. The response was framed around “better coordinating development and security,” reflecting the broader elevation of disaster prevention and resilience not only by China’s leadership but also alongside priorities linked to security and economic development.

This growing political attention comes as extreme weather becomes more frequent and costly. In mid-August, Typhoon Dolphin brought record rainfall, flooding and landslides across eastern, central and northern China, forcing more than one million people to evacuate. Climate-related losses totaled CNY 42.14 billion (US$6.1 billion) in the first half of 2026 alone and are set to increase, given that most extreme-weather events occur during the summer months. More broadly, China’s leadership is increasingly linking the growing frequency and severity of extreme weather to climate change.

Disaster response is therefore becoming more than an emergency-management issue for Beijing. As the economic and social consequences of extreme weather become harder to ignore, the leadership is increasingly linking disaster preparedness, adaptation, and resilience to climate change—and treating them as essential to protecting economic development, national security and social stability.

Subscribe now

Political Elites Elevate Disaster Response

Xi’s comments on the China-Nepal border are the latest instance of increased attention on disasters from top leadership. At an April 2026 group study session of the Political Bureau of the Chinese Communist Party (CCP) Central Committee, Xi stressed that disaster prevention, mitigation, and relief should be viewed from the strategic perspective of both high-quality development and high-level security. Notably, he also made a direct link to climate change by emphasizing that extreme weather events have become significantly more frequent and destructive.

These links to climate change have continued in subsequent meetings and publications. In July 2026, Xi chaired a committee meeting that urged all regions and relevant departments to take solid measures to ensure effective flood control, drought relief, emergency responses, and disaster mitigation. It was noted that China would continue to face more extreme weather and climate events, with both droughts and floods posing significant challenges.

Most recently, “disaster prevention, mitigation, and relief capabilities” was the headline theme of the August edition of Qiushi, the official theoretical bimonthly journal of the CCP. Its inclusion and, more so, its headlining indicate that climate is a priority for leadership. The edition led with the full text of Xi’s April speech and included several articles on the theme of extreme weather response and resilience. One of the pieces, authored by the Party Committee of the Ministry of Emergency Management, reflects Xi’s thoughts on prioritizing disaster prevention, never “lowering the guard regarding risk prevention and control” and acknowledging that China faces the “long-term, complex, and severe nature of natural disaster risks.”

When viewed together, these interventions suggest that disaster response is being elevated from a technical responsibility of emergency-management agencies into a strategic concern of the Party-state. Climate change is increasingly part of this framing because it compounds the physical risks that Beijing already sees as threats to economic development and national security. This also places greater emphasis on climate adaptation, which means taking action to prepare for and adjust to current and expected climate impacts, including more frequent and severe extreme weather events.

From Beijing’s Priorities to Local Action

The importance of responding to climate impacts and continuing to develop better disaster response is not only seen in Xi’s rhetoric but also in national-level policies. China’s “15th Five-Year Plan (2026–2030) for National Response to Climate Change”, released in July, calls for stronger climate-risk assessments and responses to extreme weather, alongside measures to build more resilient cities, food systems and health infrastructure.

Following suit, the Ministry of Natural Resources in August similarly shifted its approach to geological disasters, from a focus on individual hazard sites toward the management broader risk zones. Its “National Geological Disaster Prevention and Control ‘15th Five-Year Plan’ Implementation Plan” includes relocating around 50,000 households from high-risk areas, more than 4,200 engineering projects, expanded monitoring and early-warning systems, and greater use of big data and AI to improve disaster management coordination between national and provincial authorities.

China’s provinces and cities retain significant responsibility for translating national priorities into local action, reflecting the “一城一策” or “one city, one policy” approach, which recognizes differences in local topography and climate risks. In 2024, China’s Ministry of Ecology and Environment (MEE) deepened pilot projects aimed at building climate resilience in 39 cities while also creating unique provincial level action plans for adaptation that include sponge-city measures and risk-assessments. However, continued funding and capacity support is needed to ensure local governments do not have to choose between longer-term adaptation projects and immediate development and economic growth needs.

At the local level, citizens are increasingly engaging with climate impacts. On Douyin and Xiaohongshu, two of China’s most popular social media platforms, users share firsthand accounts of their experiences with extreme weather. During a July 2025 heatwave in Chengdu, Xiaohongshu users posted photos and videos showing parts of the city’s metro being used as a refuge from extreme heat and humidity. Research shows that during typhoons, climate impacts posting on Chinese social media rose from 41.74% to 48.97%. While this does not mean that every individual disaster is explicitly attributed to climate change, it suggests that extreme weather events can heighten public awareness of their connection to climate change.

This growing visibility of climate impacts is also reflected in public opinion: a 2024 UNDP survey found that 68% of people in China said climate change was affecting major decisions in their lives, while 39% said they were more worried about climate change than the previous year in the context of changing extreme weather. 71% also supported greater protection for people vulnerable to extreme weather.

What Does This Mean for China’s Domestic Governance and Global Diplomacy?

Domestically, disaster response, adaptation, and resilience are increasingly being framed as matters of national security and economic resilience, aligning with Xi’s broader emphasis on high-quality development, technological innovation, and social stability. The political question is therefore shifting from whether Beijing recognizes these risks to whether the Party-state can translate this increasingly clear top-level priority into adequately funded and coordinated action at the local level, where frontline implementation will take place.

For energy policy, this will also have implications as climate resilience becomes increasingly intertwined with the need to protect power infrastructure, manage rising electricity demand during extreme heat, and ensure reliable energy supplies amid droughts, floods, and other disruptions. As China’s economy becomes more electrified, the resilience of its power system will become an increasingly important component of energy security.

Diplomatically, disaster resilience offers China an opportunity to position itself as a provider of practical climate solutions while advancing Xi’s broader goals around global governance. At COP29, China launched the Action Plan on Early Warning for Climate Change Adaptation (2025–2027), while Vice Premier Ding Xuexiang announced a South-South cooperation initiative providing developing countries with meteorological observation equipment, early-warning systems, and technical training.

China’s pioneering of early-warning systems, like MAZU (Multi-hazard, Alert, Zero-gap and Universal), and its applications of AI, drones, and autonomous technologies for weather forecasting and disaster response allow Beijing to align several interests, including strengthening South-South cooperation, expanding the role of Chinese technology, and building diplomatic influence in an area where demand is growing rapidly. These technologies also give Beijing a practical way to demonstrate the benefits of Chinese innovation.

Beijing clearly understands the need to adapt and the implications on national security and economic development implications. The next challenge is turning this political recognition into sustained investment and local capacity—and, in parallel, determining how the technologies developed to manage these risks at home can support China’s wider ambitions abroad.

I visited Beijing for several weeks in June and July for a series of intensive dialogues with Chinese policymakers, advisors, corporate leaders, think tank experts, journalists, and foreign diplomats. My meetings covered a broad agenda, including China’s energy and climate strategies and industrial policy, its trade and economic relations with other nations, and its evolving view of China’s international responsibilities. Drawing on these exchanges, I offer five observations.

1. China’s economic strengths and weaknesses are strikingly visible in one sector: electric vehicles (EVs). My schedule of six to seven meetings a day in Beijing meant numerous Didi rides. For a clean tech analyst, those trips also gave me ample opportunity to observe China’s EV ecosystem up close. The variety of EV brands, positive user experience, and attractive technologies on offer in China are all well documented. Many of Beijing’s shopping malls have been repurposed as EV showrooms, showcasing China’s industrial prowess. But the sheer prominence of the sector also reflects the economy’s deep imbalances.

Following the bursting of the real estate bubble, China has struggled to identify new growth engines and appears increasingly reliant on a narrow set of sectors, including EVs, to sustain growth. But even that momentum has slowed. Passenger EV sales decelerated in June on a year-over-year basis. The service economy also remains depressed, and economic pessimism has become widespread. A contact at a leading carbon consultancy reported that most of their government contracts have dried up due to local budget constraints and that the firm has shrunk to one-third of the size it was five years ago. The NGO sector, meanwhile, is entering a period of contraction. Many organizations have frozen hiring after two decades of rapid expansion, though this retrenchment appears to be driven less by the broader economic slowdown than by funding uncertainties stemming from geopolitical tensions.

2. China’s industrial ecosystem is not only more cost-competitive but increasingly greener than production in other countries. In three separate conversations, my interlocutors pointed to new developments in China’s industrial ecosystem that are reinforcing China’s advantage: new aluminum production is relocating from coastal provinces to western China in order to leverage large-scale renewables; European chemical companies that initially moved to China for cost reasons are now finding their local production less carbon-intensive; and green iron ore production, or the processing of iron ore with minimal carbon emissions, is becoming more cost-competitive in China than in Australia, thanks to renewable energy integration. These examples all tell a larger story of how abundant, cheap green energy is beginning to fundamentally reshape China’s industrial landscape. The West’s advantage in lower-carbon production is steadily eroding while Chinese analysts expect China’s competitive lead to widen further. A presentation from a leading research agency explored a critical strategic question: What are the chances that other countries will surpass China in clean tech competitiveness, whether through incremental sector development or disruptive technological breakthroughs? The agency’s conclusion—based on detailed technology-specific analysis, supply chain mapping, and market share breakdowns—was that overtaking China in the most consequential and widely deployed low-carbon technologies is unlikely. This view is shared by a battery expert, who emphasized the accumulated engineering know-how Chinese companies have built up through years of manufacturing—expertise that is difficult to transfer and almost impossible to replicate in places lacking production capacities.

3. Economic and trade tensions, most notably between Brussels and Beijing, are likely to persist and will require fresh perspectives to manage effectively. China’s international relations experts are not optimistic about the EU-China relationship, acknowledging the direct economic competition between two parties. Some have even posed a provocative question: Will China lose the EU in the same way it lost the United States? Others have countered with the reverse: Will the EU lose China? While the relationship appears to have avoided an immediate escalation following the European Council’s mid-June meeting on China, recent interactions, including PRC Commerce Minister Wang Wentao’s visit, seem to have generated little goodwill. Both sides will almost certainly dig in ahead of the next European Council meeting in October, and Brussels will pay close attention to the September Trump-Xi meeting in the United States.

The bottom line is this: the direct conflict of economic interests between China and the EU is real and will remain a feature of the relationship. If both sides insist on rhetorical confrontation, it will generate endless friction. The danger is that these tensions may ultimately lead both sides to pursue policies that work against their own interests. Clear examples already exist. In clean tech, for instance, there is a persistent tendency to overstate the economic benefits of domestic production while underappreciating other parts of the value chain. In the solar sector, for example, installation generates more jobs than manufacturing. In the auto sector, vehicle sales and leasing, as well as an increasingly broad range of business models built around EVs, also generate substantial jobs and economic value. The broader industrialization debate could also benefit from articulating the “win.” Rather than rhetorically championing diversification or domestic production, policymakers from both sides need to specify exactly how much should be reshored, to where, at what cost, at whose expense, and on what timeline. Instead of focusing primarily on industrialization “inputs”—subsidies in the U.S. context and regulations in Europe—the expected “outputs” and “outcomes” also need to be clearly articulated. For example, how many leading companies should be in Europe, or what percentage of completely “China-free” domestic production in a given sector should be expected by 2030, and what level of subsidies or policy intervention would be required to get there?

4. The Strait of Hormuz closure has both reinforced China’s energy security strategy and prompted it to evolve. According to China’s oil experts, a combination of measures has helped to drive a sharp decline in oil imports in recent months: reducing spot and long-term purchases in response to high oil prices, suspending commercial and strategic reserve purchases, releasing reserves onto the market, imposing export controls on refined products, clearing warehouse inventories, and rapidly adopting EVs. In addition, the coal chemical sector played just a minor supporting role in reducing dependence on foreign energy supplies. The prevailing sentiment in Beijing is that this comprehensive set of measures, many of them planned and executed over years, has served China well during the current global market volatility. My Didi drivers concurred, telling me that the fuel cost of an EV is at most one-fifth that of a gasoline vehicle in Beijing and sometimes even lower. Some said they would not be in the ride-hailing business at all with a conventional vehicle. Indeed, unless one drives a gasoline car or flies frequently (a fuel surcharge was imposed earlier in the war), the Iran energy shock seems rather remote in the daily life of a Chinese citizen.

The Middle East crisis has also subtly but importantly shaped China’s thinking on energy security. An oil sector executive highlighted a shift in priorities from ensuring supply security to ensuring supply stability—the former emphasizing sufficient volumes and the latter minimizing price volatility once those volumes are secured. Another important lesson is that energy security can come from not only securing supplies but also from maintaining flexibility in demand. Compared with the significant political push from the coal industry in 2022 following the Russia-Ukraine conflict, another expert observed that the Iran crisis did not seem to trigger a similar push from the coal sector. That said, Chinese officials largely maintain their existing position on new coal power plant construction, viewing coal power as an additional source of grid flexibility and pointing to continued growth in electricity demand as industries increasingly electrify. There is little indication that they intend to curb new construction. The coal chemical sector also emerges as a clear beneficiary, now bolstered by strategic considerations that diminish the force of environmental and economic counterarguments.

5. China’s sense of international responsibility is evolving, offering new structural opportunities for environmental advocacy. Chinese experts noted that recent policy documents have taken an important step, vowing to “firmly” take the lead (从“引领”到“坚定引领”) in multilateral governance—a phrase first used in 2017 during the first Trump administration and a sign of the leadership’s resolve to take a more proactive role. China’s strategic community appears to be moving ahead of its climate community in thinking about the country’s global role. In their view, China’s expanding global interests increasingly require a more proactive international posture. This position will gradually shape downstream policy domains, including climate change, making climate responsibility an extension of China’s broader global strategy. Even conversations within the climate policy circles offer revealing insights into China’s perception of its future global role. The Chinese climate envoy and policy advisors all expressed hope that Washington would return to the UN Framework Convention and the Paris Agreement. This desire should not be dismissed as merely rhetorical. One could even ask: In an increasingly competitive relationship, why would China want the United States back in the multilateral climate system, only to face more pressure from Washington? The answer lies in a deeper shift in how China views its international climate position, particularly its cost-benefit calculus. As the world’s dominant producer of clean energy technologies, Beijing increasingly sees ambitious global climate action as advancing its own economic interests. Sustained American absence, in China’s view, weakens the global demand needed to support that commercial opportunity.

Given that AI governance could emerge as a topic ahead of the anticipated September 24 Trump-Xi summit, we bring together on-the-ground observations from three CCA contributors who traveled to Shanghai for the 2026 World Artificial Intelligence Conference (WAIC): Alvin Wang Graylin, Paul Triolo, and Ran Guo. Their contributions examine how China’s AI ecosystem is evolving across compute, governance, safety, commercialization, and data. Kristy Loke, a research fellow with the Machine Learning Alignment & Theory Scholars (MATS) program who studies China’s AI governance, adds a broader strategic lens by examining Xi Jinping’s articulation of an open AI strategy and its implications for U.S.-China technology competition. Taken together, these observations offer a corrective to a chip-centric view of U.S.-China AI competition. Compute constraints are real, but the conference revealed an ecosystem adapting around them. The governing question for U.S.-China technology policy is therefore how export controls interact with China’s increasingly durable adaptations.

For a broader take on the themes across this issue, read Lizzi C. Lee’s Substack post: When the Constraint Becomes the Strategy.

In this issue:

1. Alvin Wang Graylin — Compute constraints, World Artificial Intelligence Cooperation Organization (WAICO), policy signals, and data

2. Paul Triolo — Hardware integration and China’s safety-evaluation ecosystem

3. Kristy Loke — Openness as China’s strategic narrative

4. Ran Guo — Data governance and the economics of health AI

Terms readers need

FLOPs and compute scarcity: FLOPs, or floating-point operations, are a measure of computing work. In this issue, compute scarcity refers to limited access to leading accelerators for training, inference, and safety evaluation.

Blackwell and the Atlas 950 SuperPoD: Blackwell is Nvidia’s advanced accelerator architecture. Huawei’s Atlas 950 SuperPoD is a system-level cluster designed to link as many as 8,192 Ascend 950 NPUs into a single large, logical computing system.

Open-weight: An open-weight model is one whose trained parameters are released for others to download, modify, and deploy, even when the full training data or development process is not publicly available.

1. Alvin Wang Graylin: Compute adaptation, institutions, and data

By Alvin Wang Graylin, Non-Resident Honorary Senior Fellow on Technology, Asia Society Policy Institute’s Center for China Analysis

Compute constraints, WAICO, open-weight economics, safety priorities, robotics, data, and engagement options.

Key takeaways

• Compute scarcity is now a managed condition rather than a crisis. China’s labs still want more high-end silicon, but they have restructured around the constraint. The cost is measured not only in floating-point operations (FLOPs), a standard measure of computing work, but also in engineering time spent adapting models to domestically produced AI chips.

• Export controls and entity-list designations—not tariffs—are now the main grievance. Multiple interlocutors said tariffs have receded as a concern, making delisting the most requested goodwill signal—one framed as low-cost for Washington and highly symbolic for Beijing.

• Chinese frontier labs generally treat safety as a compliance obligation rather than an independent research priority, and several interlocutors argued that U.S. compute restrictions can undermine safety by leaving labs with fewer resources for testing and evaluation.

• At the regulatory level, Chinese authorities remain focused primarily on information security and content governance — ensuring that AI-generated content complies with state content rules and avoids politically sensitive topics — rather than on catastrophic misuse or the possibility that advanced AI systems could move beyond effective human control. As a result, proposals framed around such loss-of-control scenarios are unlikely to gain traction; those centered on shared mitigation of bad actors in cyber, bio, and chemical domains may prove more viable.

• At the institutional level, WAICO was the conference’s main development: 29 founding signatories, a Shanghai headquarters, no major Western democracy, and a keynote from Xi presenting it as a milestone. It is an infrastructure-and-capacity initiative aimed at the Global South, running alongside rather than within the UN track convened in Geneva two weeks earlier. Xi again emphasized open source and AI as a public good.

• At the same time, the open-weight consensus may be less durable than Chinese labs believe. Their leaders reported no government pressure to close models, even as public reporting indicated that Beijing was considering export restrictions on advanced domestic models.

Compute: real constraint, real workarounds

The clearest evidence of adaptation to China’s compute constraints appears at the compute layer. Practitioners repeatedly described Huawei’s accelerators as roughly two generations behind leading international chips, with one calling them “less than H100.” Frontier model training therefore still relies largely on international Blackwell-class hardware accessed through offshore partners. At the same time, the engineering work required to adapt and optimize models for domestic silicon adds another burden to China’s compute deficit.

The public record reinforces this picture. Huawei used WAIC to debut the Atlas 950 SuperPoD, an 8,192-chip system. The strategy is clear: compensate for weaker individual dies through system-level scale-up, interconnect, and cluster engineering. This is an expensive but viable route to higher training throughput, though it is considerably less effective at reducing inference costs per token.

The debate cuts both ways. Restriction advocates argue that the two-generation gap and domestic tuning overhead are precisely the intended effects, and congressional projections of indigenous H200-equivalent capability in 2028 suggest the policy is buying time. Critics counter that offshore training, scale-up architectures, and algorithmic efficiency are the adaptations the controls were meant to prevent—and that successful workarounds become permanent while U.S. export controls remain revisable.

My assessment is that the controls impose measurable costs and are not being cleanly circumvented at the frontier. But adaptation is now institutionalized, reducing the return on further tightening while raising the costs to bilateral trust and revenue earned by U.S. firms. The policy is working narrowly while losing ground more broadly.

Model layer: economics and openness

At the model layer, economics and openness are also reshaping the market. Domestic inference monetization remains weak, except in video generation, where Chinese consumers and businesses demonstrably pay. Text and general-assistant inference generate little meaningful domestic revenue. As a result, international inference demand has become the growth engine, with an increasing share of Chinese lab revenue coming from outside China. This matters for assessing the commercial effects of restricting Chinese model access abroad.

DeepSeek, meanwhile, is now partially state-owned and has no conventional business model. Its valuation rests on brand equity as a national champion and on the expectation that the state will underwrite its long-term success. It intends to remain open-weight and is moving toward multimodality on the view that multimodality is a prerequisite for AGI.

Across Chinese labs, AGI is defined instrumentally as “smart enough to do human work,” rather than as a conscious or godlike system. It is viewed as one long-term goal among several, not a singular technological end point. This difference in framing may fuel mutual misperceptions between Chinese practitioners and parts of the U.S. AI debate.

These pressures have also narrowed the field sharply. A year ago, the expo floor featured more than 100 models; that number has since fallen dramatically. Regional champions remain, but consolidation is well advanced.

Openness, however, appears to be driven more by commercial than political considerations. Lab leaders said they faced no government pressure to close weights and characterized Alibaba’s more closed posture on certain models as a business decision rather than a technical or political one.

Reuters and the Financial Times reported in July that Chinese authorities were considering tightening export controls on advanced AI models and semiconductor technology, following earlier meetings with major tech firms about restricting overseas access to China’s most capable models, including those not yet released. Either the labs are not yet in the loop, or the policy is contested internally, or the restriction under consideration is narrower than a general move to closed weights. All three readings have very different implications for the open-versus-closed debate, making this the most important issue to track over the next quarter.

Safety and governance: what Chinese actors actually prioritize

These technical and commercial shifts also shape how Chinese actors approach safety and governance. Contacts close to China’s regulatory apparatus described the operative definition of “safety” as security and information control. Attention to CBRNE misuse is comparatively thin, and loss-of-control risk remains largely absent from lab priorities, though it is gaining attention in some safety institutions. At the lab level, developers follow government rules and already view existing constraints as burdensome. Several argued that if safety were a genuine U.S. priority, compute availability would increase rather than decrease because resource-constrained labs are more likely to cut corners. Cooperation is possible, but it depends on framing.

Even within these constraints, there are openings for cooperation. An International Atomic Energy Agency (IAEA)-style international standard for safety evaluation, potentially linked to the UN, drew unprompted support from multiple interlocutors—the most concrete multilateral opening I encountered. The clearest practical proposal emerged from a session on AGI cooperation, which made the week’s strongest case for governance interfaces rather than legal convergence. Drawing on private international law, the argument was that shared rulebooks are unnecessary if interfaces, adapters, and conflict rules exist. Cooperation without trust could rest on shared vocabularies and protocols, with the 1972 Incidents at Sea Agreement as a precedent. Mutual recognition and shared test suites could provide enforcement, producing a common incident taxonomy, reporting schema, and triage thresholds.

Yet the trust deficit remains the binding constraint. Interlocutors were willing to cooperate but doubted that the United States would honor agreements. A recurring observation was that AI diplomacy lacks a Kissinger-like figure with standing on both sides.

WAICO and the institutional architecture

The same emphasis on practical cooperation and capacity building is visible in the emerging institutional role for the World Artificial Intelligence Cooperation Organization (WAICO). Twenty-nine countries signed the founding agreement on July 16, one day before WAIC opened. Wang Yi signed for China, and Guterres attended. Headquartered in Shanghai, WAICO is an independent intergovernmental body invoking UN Charter principles. Reported founding members include Russia, Pakistan, Indonesia, Kazakhstan, Laos, Belarus, Serbia, Brazil, Cuba, and Venezuela. In his first WAIC keynote in the summit’s nine-year history, Xi announced 5,000 AI training opportunities for participants from developing countries over five years, along with cooperation centers with ASEAN and the African Union.

At the same time, WAICO is emerging alongside the UN track. The first UN Global Dialogue on AI Governance convened in Geneva on July 6–7, marking the first dedicated meeting of all UN members on the subject. Guterres advanced a Global Fund for AI covering skills, data, and affordable compute, and said more than 20 member states had nominated centers for a UN-supported Global Network for AI Capacity Building. Among more than 1,500 submissions, most non-government groups ranked safety first, while governments prioritized capacity building.

This overlap between the UN Global Dialogue on AI Governance and WAICO supports two competing interpretations. The competitive view is that WAICO is a parallel institution designed to set standards and lock in dependencies before the UN track produces anything binding; the absence of major Western democracies is the point. The complementary view is that WAICO’s agenda—capacity building, training, and compute and data access for developing states—closely matches the proposed UN Global Fund, creating an opportunity for integration rather than a fork.

My assessment, then, is that the competitive reading is right about intent and the complementary reading about substance. That argues for engagement rather than boycott. Ignoring WAICO would cede the Global South AI capacity agenda just as the UN has validated it as a priority. Observer status, technical participation in evaluation standards, and interoperability between WAICO programs and the proposed UN fund are lower-cost alternatives.

Subnational industrial policy

Below the international level, China’s AI expansion is also being driven by local industrial policy. Local governments provide office space, fast-tracked logistics, limited compute and telecom resources, and access to city and provincial leaders—the last of which is often described as the most valuable input. Within this system, regional leaders are consistently assessed on four metrics: job creation, business and tax-base growth, innovation output, and social stability. Every subnational AI policy choice maps back to these priorities.

Yet this support creates its own tension. Both central and local governments are seeking startup equity, while firms planning international expansion often resist, recognizing that state ownership can become a liability abroad—an underappreciated tension in China’s AI capital structure. At the same time, municipal incubators and AI education programs are proliferating rapidly.

Robotics: involution now, consolidation expected in 12 to 24 months

Local support is especially visible in robotics, where rapid expansion has brought both scale and intense competition. The expo featured more than 200 robotics firms, and participants openly described the sector as being in an involution phase, with consolidation expected within one to two years.

Unitree offers an important reference point. It is fully vertically integrated in component manufacturing, although roughly 80 percent of its assembly work remains manual. The company is trying to automate, but the current process, while archaic, is adequate at present production volumes. It expects costs to fall by around 80 percent over time. Unitree is also the only Chinese player currently able to ship globally with support infrastructure and partners in place, including parts replacement for 1.5 to 2 years. Even so, its 2025 volumes remain relatively small: approximately 30,000 to 40,000 quadruped robots, 5,000 to 10,000 full humanoids, and 5,000 to 10,000 non-legged humanoids. Its software and training work also remains limited, and leadership was reportedly skeptical of humanoids until relatively recently.

More broadly, the hardware and supply-chain advantage is real, durable, and underestimated in Washington. The genuine constraint is software, training, and manipulation policy—the layer most exposed to compute restrictions. Chinese embodied AI capability should therefore be measured by software progress, not unit shipments.

Beyond humanoids, non-humanoid service robots are already deployed commercially, and in major cities more than half of the vehicles on the road are electric. The physical-world deployment base is further ahead than many U.S. visitors expect.

Data: the most underrated asymmetry

Yet hardware and robotics are only part of the structural picture; data may be an even more consequential asymmetry. On healthcare access, one firm described free access to data covering roughly 300 million patients annually, spanning more than 4,000 test types and complete patient records. There is no analogous access regime in the United States or the European Union.

The capabilities presented were also substantial. A fellow of the Chinese Academy of Engineering and the president of Ruijin Hospital described findings from a 15-year cohort study of 150,000 people, including a model that predicted major adverse cardiac events over 10 years with reasonably strong accuracy (AUC of approximately 0.78), and a 25-protein panel that predicted 10-year mortality with very strong discriminatory performance (AUC of approximately 0.90).

The business model may also be evolving toward a “data dividend” model. Under this vision, hospital care would trend toward being free or paid for through data, while hospitals commercialize the resulting models and shift medicine from reactive treatment to prediction, prevention, and early intervention.

The compute debate has crowded out the data question. In health data, China’s advantage is structural, growing, and beyond the reach of export controls. It also strengthens the case for a cooperative global data pool: cross-site validation would benefit both sides scientifically without costing either side its strategic position.

Capital markets and exit paths

Turning these technical advantages into durable businesses, however, depends on financing and viable exit paths. The STAR Market (科创板) functions as the domestic exit venue and permits listing without profitability, although multiples remain well below international comparables.

That domestic route is further constrained by Beijing’s reluctance to see its leading firms list internationally, compressing the exit menu.

Hong Kong therefore retains its role as the financial bridge to global capital and is positioning itself as a compliant AI sandbox with an East-West standards role.

One indication of this role is Hong Kong’s deployment of HKPilot, a government document-processing copilot built on the locally developed HKGAI LLM. By November 2024, more than 800 officers across over 20 bureaus and departments were participating, and by April 2025 the trial had expanded to more than 70 government departments. HKGAI subsequently released V3 of its model in June 2026.

What Beijing is actually asking for

Taken together, these constraints and adaptations help explain the priorities Chinese interlocutors raised in discussions with the United States, ranked here by frequency and intensity.

The most frequent request centered around the removal of firms from sanctions, export-control, and risk lists. Alibaba and Tencent are lobbying against designations they consider unjustified, and delisting is seen as a cheap, legible signal of good faith.

Access to TSMC was presented as a more achievable request than access to EUV tools. Chinese contacts assessed domestically usable EUV as roughly five years away.

Interlocutors also expressed willingness to co-invest in the United States and in Global South AI infrastructure, preferring a new brand distinct from both the Belt and Road and the Marshall Plan.

They also raised reciprocal restraint on competitive practices. Beijing signaled willingness to push domestic firms to moderate the hyper-competitive export behavior they have been trained into domestically, acknowledging that these practices are alarming firms in every market they enter.

Huawei, however, was not part of the request of removal from the sanctions, export-control, and risk lists. The company has internally accepted its designation as permanent and does not expect that status to change.

Implications and judgment

The case for taking these openings seriously is that the requests are specific, ranked, and internally consistent. Delisting a small number of non-sensitive firms is reversible and verifiable, while interest in IAEA-style evaluation standards and mutual test recognition is technically substantive. The Global South co-investment offer aligns with WAICO’s program, and restraint on aggressive export practices would address concerns shared across the United States, Europe, Japan, Korea, and India.

The skeptical case, however, is that each opening is low-cost for Beijing but consequential for Washington. Delisting transfers real capability in exchange for goodwill. China’s safety agenda remains centered on information control, ensuring that AI-generated content complies with state rules and avoids politically sensitive topics, and could legitimize content-governance norms the United States opposes. WAICO embeds Chinese standard-setting under multilateral cover, while possible Chinese controls on model exports suggest Beijing may adopt the same playbook centered on restrictions it condemns.

My judgment is that the skeptical case is right about motive but wrong about consequence. The test is not whether an arrangement is costless to Beijing, but whether it improves on the status quo and creates a new verification surface. A conditional pilot delisting, a shared incident vocabulary, and a jointly recognized evaluation suite all meet that test without requiring trust. The Incidents at Sea precedent matters precisely because it was concluded between adversaries who expected to remain adversaries.

The more fundamental objection, therefore, is that the United States lacks both an interlocutor structure capable of executing such a technical arrangement and a domestic political coalition that could sustain it. That is the first gap to address.

What to watch over the next 90 days

1. WAICO’s first substantive work program and whether any OECD member seeks observer status.

2. Whether the UN Global Fund for AI recommendation to the General Assembly creates a channel that intersects WAICO’s capacity-building agenda.

3. BIS action on the AI diffusion rule replacement, targeted for end of fiscal year, and the fate of the Banks amendment—the AI OVERWATCH Act, which would codify tighter controls on advanced AI-chip exports to countries of concern—in the fiscal year 2027 National Defense Authorization Act.

4. Any movement on entity list removals, which would be the clearest available evidence that the delisting channel is live.

5. Robotics consolidation, including the first significant failure or acquisition among the 200-plus expo participants.

Additional observations

Apple is allegedly using Qwen models in Apple Intelligence, a claim partially corroborated by prior public reporting on the China deployment, although the current scope still requires confirmation. Tencent is reportedly acquiring Manus at a price above a competing Meta offer.

2. Paul Triolo: Hardware adaptation and China’s safety-evaluation ecosystem

By Paul Triolo, Non-Resident Honorary Senior Fellow on Technology, Asia Society Policy Institute’s Center for China Analysis

Advances across China’s AI hardware stack and the emergence of a distributed, specialized network of safety and evaluation institutions.

The 2026 WAIC in Shanghai highlighted major advances across China’s AI stack since the 2025 edition, which I also attended. President Xi Jinping’s opening speech framed the conference around two main themes: the benefits of AI diffusion, including to the Global South through the new 29-country World AI Cooperation Organization (WAICO), and the need to keep AI safe and secure. Xi’s safety comments “could have been written by me,” a leading AI safety researcher quipped during a discussion on the margins of the conference.

The conference came as U.S.-China competition in AI hardware and models reached new highs, while the release of Anthropic’s Mythos model and debate over the released, safeguarded Fable 5 counterpart intensified U.S. questions about how to gate frontier releases. Soon after, the OpenAI GPT 6.0 model’s escape from its sandbox and attack on Hugging Face underscored the need for a regulatory regime around advanced models. Moonshot’s release of Kimi K3 just before WAIC also fueled U.S. debate over Chinese open-weight models. These issues ran through public panels, closed-door meetings, and sidebar conversations near the main venues.

What stood out during my tour of the exhibition was the sharp increase in the number of Chinese AI hardware developers offering GPU clusters for training and inference. Moore Threads, Biren, Enflame, MetaX, and Huawei displayed larger GPU or Ascend NPU arrays, typically labeled as tested or “ready” for leading Chinese models from DeepSeek, Moonshot, Zhipu, Alibaba, or MiniMax. Tight integration between hardware and model software was a central theme.

This does not mean all Chinese model developers are training advanced models on domestic hardware. Moonshot has allegedly accessed Nvidia Blackwell compute in Southeast Asia, while Zhipu has trained a version of GLM-5.2 on Huawei’s Ascend architecture. The Atlas 950 SuperPoD was displayed for the first time, and a Huawei engineer walked me through its technical design. Huawei’s system-level answer to weaker individual accelerators links up to 8,192 Ascend 950 NPUs through UnifiedBus, unified memory addressing, and high aggregate bandwidth, allowing the installation to behave more like one giant logical computer than a conventional GPU cluster. The architecture should be especially useful for communication-heavy mixture-of-experts (MoE) training and high-volume inference. Its earliest users, however, will likely be Huawei/Pangu, state-backed clouds, telecom operators, SOEs, and laboratories that need a fully domestic stack—not independent frontier labs with dependable Nvidia access.

Assuming Moonshot can obtain thousands of Blackwells, it will likely keep flagship pretraining on Nvidia, because CUDA maturity, proven scaling, stronger per-device performance, and researcher productivity outweigh Ascend’s strategic appeal. Moonshot and Alibaba would move workloads gradually—starting with inference, fine-tuning, reinforcement learning, overflow capacity, and redundant training runs—and shift major pretraining only once the 950 demonstrates sustained utilization and reliability, or Nvidia supply and regulatory risks become unacceptable. Zhipu can support a serious Ascend porting and capacity-reservation program after recent capital raises, but renting Huawei or state-cloud capacity and dual-porting selected models is more plausible than operating an entire 160-cabinet SuperPoD. The full Atlas 950 is not scheduled until the fourth quarter of 2026; once leading developers have tested it, however, it could be a game changer.

The broader outlook for Chinese GPU makers is therefore improving. SMIC is likely to expand advanced-node capacity in Shanghai, with the government working alongside SMIC, Huawei, and other chip designers to allocate it. By year-end, more domestic GPU clusters and Huawei 950 SuperPoDs should be available for training and inference. Chinese developers will have greater access to advanced Nvidia hardware outside China, more domestic training capacity, and more inference compute to commercialize models such as Kimi K3. WAIC also underscored the push for stronger business models as firms list or prepare to list in Shanghai and Hong Kong, raise capital, and expand capacity. The CXMT IPO will channel additional capital into expansion, including high-bandwidth memory. Business models, markets, capex, and flexibility were hallmarks of the exhibition.

AI Safety, major evolution in thinking since WAIC 2025

The panels, closed-door meetings, and sidebar conversations on AI safety also showed how much thinking has evolved since the Paris AI Action Summit and last year’s WAIC. Xi’s reference to AI safety was groundbreaking and signaled that China’s fast-growing safety community should now be taken seriously. Western debate has traditionally focused on catastrophic risks, loss of control, and misalignment. In China, anquan （安全; safety）has more often referred to public safety and social stability, with less attention to loss-of-control or national-security risks involving cyber operations or CBRN capabilities. WAIC suggested this is beginning to change.

Rather than creating a single equivalent of the UK AI Security Institute or the U.S. Center for AI Standards and Innovation, Beijing is assembling a distributed network of frontier laboratories, governance institutes, standards organizations, and model evaluators. WAIC suggested that this ecosystem is maturing into a potentially important counterpart for international safety cooperation. China’s AI safety community has come a long way.

For much of the past two years, discussions about China’s approach to AI safety have focused on what appeared to be missing. Compared with the UKAISI or CAISI, China seemed to lack a single institution responsible for evaluating frontier AI systems before deployment. Safety research existed, regulators had established security assessment procedures for generative AI models, and leading universities published internationally on interpretability and alignment, but these activities appeared fragmented and often disconnected. The dominant question among foreign observers became: Where is China’s AI Safety Institute? As I noted last year, the emergence of the China AI Safety and Development Association (CnAISDA) at the Paris AI Action Summit was an important milestone in this process.

But more than one year later, discussions at the 2026 World AI Conference suggested that this question may have been the wrong one all along. Rather than unveiling a centralized institution modeled on Western counterparts, the CnAISDA was a network of organizations, and the 2026 WAIC highlighted a growing network of organizations that collectively perform increasingly important functions. Throughout the conference, in addition to CnAISDA, the Shanghai AI Laboratory, Tsinghua University’s Institute for AI International Governance, researchers associated with the Beijing Institute of AI Safety and Governance, standards bodies including CAICT and TC260, and major frontier AI developers appeared together in technical workshops, policy discussions and international dialogues. The picture that emerged was not one of institutional duplication, but of an increasingly specialized ecosystem in which different organizations contribute distinct capabilities.

This distributed model is not an accident. It reflects China’s broader approach to governing strategic technologies. Rather than concentrating authority in a single independent agency, Beijing frequently relies on overlapping institutions that combine research, industrial policy, standards development, and government coordination. The same pattern can be seen in semiconductors, quantum technologies, and biotechnology. AI safety increasingly appears to be following a similar trajectory.

Shanghai AI Laboratory: Building China’s Technical Evaluation Engine

At the WAIC in Shanghai, one of the most interesting presentations came from Professor Zhou Bowen, the Director of the Shanghai AI Laboratory (SAIL). SAIL has quietly emerged as perhaps the country’s most technically capable institution for evaluating frontier AI systems. Although widely known internationally for the InternLM family of open-weight language models, InternVL multimodal systems, and AI for Science initiatives that I have documented in my Substack, the laboratory’s significance extends well beyond model development. By training frontier-scale models itself, SAIL has developed precisely the engineering expertise required to evaluate increasingly capable AI systems.

This distinction matters. Evaluating frontier models involves far more than prompting public APIs or measuring benchmark performance. Researchers require access to distributed training infrastructure, inference pipelines, model checkpoints, activation data, and internal architectures. They need to understand how models behave under different decoding strategies, how safety fine-tuning affects downstream capabilities, and how evaluation benchmarks themselves can become contaminated. Institutions that build frontier models inevitably acquire expertise that cannot easily be replicated through external testing alone.

SAIL has complemented this expertise by developing one of China’s most sophisticated model evaluation infrastructures. OpenCompass has evolved into a comprehensive evaluation framework covering hundreds of language and multimodal benchmarks across both open and proprietary models. Rather than treating evaluation as a one-off benchmarking exercise, OpenCompass increasingly resembles shared scientific infrastructure, enabling reproducible comparisons across successive generations of frontier models. Its importance within China’s AI ecosystem parallels the role played by platforms such as HELM or LM Arena internationally, while extending more deeply into Chinese-language evaluation and multimodal systems.

The laboratory has also expanded its explicit focus on AI safety. Through its AI45 initiative, SAIL now conducts research spanning trustworthy AI, embodied AI safety, interpretability, AI for Science safety, and broader questions surrounding advanced AI systems. While much of this work remains technically oriented, it reflects an important evolution. Safety is no longer treated primarily as content moderation or compliance with existing regulations. Instead, SAIL frames safety increasingly as an engineering challenge requiring systematic evaluation throughout the model development lifecycle.

Zhou’s presentation at WAIC 2026 demonstrated how quickly China’s technical evaluation capabilities are maturing. Rather than focusing on AI governance principles or regulatory frameworks, Zhou emphasized the engineering challenge of evaluating increasingly capable frontier models and AI agents throughout the model lifecycle. His framework integrated capability assessment, safety evaluation, agent behavior, tool use, and continuous testing into a unified evaluation pipeline, reflecting the perspective of a frontier model developer rather than a policymaker. This approach is particularly significant because it suggests that organizations such as SAIL are developing the technical infrastructure—including evaluation platforms such as OpenCompass, benchmark engineering and large-scale testing methodologies—needed to systematically characterize the capabilities and failure modes of frontier models. In many respects, SAIL appears to be evolving into China’s closest analogue to an engineering-focused evaluation laboratory, where safety is treated as an extension of model development rather than simply a regulatory compliance exercise. The emphasis is on building scalable evaluation systems that can keep pace with rapidly advancing foundation models and increasingly autonomous AI agents.

By contrast, the work emerging from Tsinghua University’s Institute for AI International Governance (I-AIIG) and Zeng Yi’s Beijing Institute of AI Safety and Governance begins from a different premise: defining what constitutes frontier AI risk and how those risks should be incorporated into national and international governance frameworks. Tsinghua has focused on developing international evaluation standards, safety capability maturity models and mechanisms for cross-border cooperation, while Zeng Yi’s group has pushed China’s frontier risk research toward questions of agentic autonomy, AI for Science, embodied AI, catastrophic and even existential risks through initiatives such as ForesightSafety Bench.

Their work increasingly resembles that of organizations like CAISI or the International AI Safety Report community, emphasizing comprehensive risk taxonomies and scientifically grounded evaluation methodologies rather than benchmark engineering alone. Taken together, these two communities illustrate an emerging division of labor within China’s AI safety ecosystem: Shanghai is building much of the technical machinery needed to evaluate frontier models, while Beijing is defining the conceptual frameworks, governance architecture and international standards that determine which risks should be evaluated and why. The challenge now is institutional—integrating these complementary strengths into a trusted national capability for independent frontier model evaluation, comparable to the role that organizations such as METR, Apollo Research and the UK AI Security Institute are beginning to play in Western AI governance. During my panel presentation on prospects for international cooperation, I mentioned the proposal from Google DeepMind CEO Demis Hassabis for a self-regulatory organization (SRO) in the United States that over time could be a concept embraced more globally, and I also stressed the need for independent third-party evaluation organizations as part of an emerging consensus in the United States and the broader AI safety community.

For China, however, WAIC also highlighted several challenges in this area. SAIL’s strengths, for example, also reveal its limitations. Despite its technical sophistication, it remains simultaneously a frontier model developer, a government-supported research institution, a commercial operation, and a potential evaluator of competing models. There is little public evidence that it possesses statutory authority to demand pre-deployment access to proprietary models developed by Alibaba, Tencent, DeepSeek, Zhipu, or Baidu, or that its evaluation findings automatically influence regulatory approval. In other words, SAIL has accumulated much of the technical capacity associated with a frontier model evaluator without yet occupying the institutional role of an independent national evaluation authority.

Perhaps the most significant implication of WAIC 2026 is that China is no longer asking whether frontier AI requires dedicated safety institutions. That debate has largely been settled. The more important questions now concern how responsibilities should be divided across an increasingly diverse institutional landscape, how evaluation methodologies should evolve alongside rapidly advancing frontier models, and how China’s emerging ecosystem should engage with counterparts abroad.

For international AI governance, this evolution matters enormously. As the United States, United Kingdom, Japan, Singapore and others continue building their own frontier model evaluation capabilities, China is constructing a parallel—though institutionally distinct—architecture. The differences are real, particularly regarding independence, transparency and regulatory authority. Yet the underlying technical challenges are remarkably similar. Evaluating increasingly autonomous, multimodal and scientifically capable AI systems will require common methodologies, shared benchmarks and ongoing dialogue across national boundaries. Some of this will begin happening when the United States and China meet in September for the first time to discuss real frontier model guardrails and regulation. This will not be easy, but Chinese organizations now have much more to bring to the table than was the case in Paris in early 2025 or the WAIC in July 2025.

Competition over frontier AI capabilities is likely to intensify. But if WAIC 2026 demonstrated anything, it is that competition need not preclude cooperation on the science of evaluation itself. Indeed, the emergence of increasingly sophisticated AI safety ecosystems on both sides may finally provide the institutional counterparts necessary for sustained technical engagement on one of the defining governance challenges of the AI era.

3. Kristy Loke: Openness as China’s strategic narrative

By Kristy Loke, Research Fellow, Machine Learning Alignment & Theory Scholars (MATS) program

Beijing’s open-AI strategy and its implications for competition, governance, ecosystem power, and selective U.S.-China coordination.

Openness as a long-term commitment

A popular question among China watchers is: what does China want from AI? At the World Artificial Intelligence Conference (WAIC), President Xi Jinping offered his clearest answer so far. China wants to partake in, be a leader in, and share the benefits of AI – and it wants to do so through a strategy of openness. Xi summarized China’s global AI contributions in a four‑point framework – deepening the strategy of openness and promoting win–win development, improving risk awareness and ensuring safety, security, and controllability of AI, encouraging multiculturalism, and promoting consensus-based global governance – with openness sitting at the heart of all four.

Whether future Chinese AI takes the form of open‑weight AI releases, which has become the focus of recent debates, is almost beside the point; the commitment is more fundamental. Openness is a long‑term orientation: not just via open‑weight models, but by encouraging Chinese AI firms to keep sharing the know-how, infrastructure, and resources for others to build, own and deploy AI. China’s leaders believe there are enough benefits from AI to go around, and that openness is the path to diffusion and value creation. By defining its AI vision around openness, Beijing implicitly rejects an AI arms‑race framing in which a few months’ frontier AI model lead is seen as decisive.

How will China govern openness?

Analysts have often portrayed Chinese leaders and regulators as reckless, anxious, or rigid in balancing AI development and governance. Recently, this framing has resurfaced in warnings about reflexive overregulation or blanket bans on frontier open models. So far, the evidence suggests otherwise.

Since 2023, Beijing has pursued a pro‑governance and pro‑development approach. The AI‑Generated Content Interim Measures laid the groundwork for a pre‑ and post‑deployment transparency regime, with a focus on controlling prohibited, harmful, or politically sensitive AI-generated material, and were paired with a State Council message for regulators to tread lightly and smartly: grasp emerging risks, issue timely and narrow prohibitions, and avoid deterring good‑faith innovation. Within this context, models such as DeepSeek-V3, GLM-5.2 and Kimi K3 managed to emerge. Remarkably, this commitment to governance has not prevented China’s AI model sector from developing, keeping up, and at times, leading.

For open-weight AI, we should expect a similar balance: prohibiting clear misuse and unchecked capabilities, encouraging experimentation where care can be demonstrated, and continuing to explore risk tiers, pre-deployment testing, disclosure, and post-deployment monitoring. If that balance holds, the incentive for a blanket ban remains low.

From containment to an open AI strategy

China’s embrace of open innovation predates the LLM wave. Under repeated rounds of tech containment since 2018 – from Trump’s ZTE and Huawei sanctions to Biden’s tightening chip controls – Beijing was forced to rethink its tech and innovation priorities. Within weeks of the ZTE shock, Xi told leading scientists that core technologies such as semiconductors could not be “obtained, bought, or begged for,” prompting debate over which parts of the globalized chip supply chain had to be brought home. The debate ultimately landed on the side of the importance of core tech indigenization and continued links with innovation channels and sources abroad. Xi’s WAIC reference to “Chinese‑made intelligence” (中国智造) captures this ambition: to be a key exporter of AI capability by being a reliable, trustworthy and relatively affordable supplier, in turn benefiting from both business and creative ideas and innovation from abroad.

The strength of Chinese open models reinforced this open orientation. When DeepSeek-V3 was released in late 2024 as a cheaper, more manipulable alternative to proprietary U.S. systems, the U.S. stock market and global developers took notice. In line with its pragmatic strategy, Beijing and its advisers moved to further center openness in China’s AI posture at home and abroad. This is evidenced by growing legal discourses over how to best govern open-weight AI to help it develop safely and by various Qiushi pieces acknowledging open-source AI as central to China’s AI strength. Given this context, Xi’s WAIC speech is best read as the culmination of that evolution, not a sudden pivot.

Openness as China’s AI edge—and a challenge for the United States

China’s commitment to openness serves both domestic and external goals. At home, openness promotes more accessible AI and faster innovation‑led economic transformation. Abroad, openness and shared benefits are framed as an antidote to an “AI Iron Curtain,” positioning China’s open models as affordable and reliable options for developers worldwide.

By contrast, the United States is quietly starting to lose three important competitions by over‑emphasizing frontier AI lead and containment as a winning strategy, and in the process, inadvertently prioritizing tech supremacy, exclusivity, and unilateral actions. In trying to protect and widen the United States’ frontier AI edge by any means possible, the United States is also making itself less competitive on AI accessibility, cost, ecosystem pull, and soft power, thus losing both friends and influence. At home, the lack of an inclusive and distributive vision for AI also deepened economic and environmental anxiety among a growing part of the population. While it is not too late for the United States to reconsider, an AI strategy centered around openness will look quite different from the one it has pursued over the past decade.

The narrow window for U.S.–China AI coordination

Recent weeks show both the pace of progress and the sharpness of emerging risks. GLM-5.2 and Kimi K3 were announced just before WAIC. Soon after, Hugging Face disclosed a cyber incident involving OpenAI in which GLM-5.2 was used defensively after a U.S. proprietary model refused requests for cyber defense. The episode highlights cross-border entanglement and growing challenges of control and security. It also undercuts the idea that U.S.-China AI development and governance are inherently zero-sum: models from one country are already defending against vulnerabilities in another.

From WAIC, notably, Xi’s framing of AI risks as “internal” and “external” (to AI systems) overlaps with concerns in the West over frontier AI control and misalignment risks, and misuse risks, even if the language differs. The reality is that the two are facing a similar and growing set of frontier AI risks, and as the leading AI powers, have the most to lose from unsafe AI development and deployment. Only by recognizing this reality can Washington move from containment instincts to much-needed selective coordination with Beijing around such risks. The time to act is now.

4. Ran Guo: Data governance and the economics of health AI

By Ran Guo, Affiliated Researcher, Asia Society Policy Institute’s Center for China Analysis

Data governance, trusted data spaces, and the commercial logic that requires health-AI products to prove system-level savings for public insurance.

Over ten days, from July 9–19, I attended events in Beijing and Shanghai: the 2026 China Internet Conference (“Data for AI” and “AI + Health” forums), a Tongji University workshop on AI safety and security, and WAIC. Two themes stood out: data governance and AI + health.

1. Data governance: imagination, incorporation, and new practices

My strongest impression is that many practitioners, analysts, and policymakers approach AI governance through data policy, especially data use and security. China’s developmental and regulatory approach has moved from “大数据” and “数字化” (“big data” and “digitalization”) toward “数智化” (digital intelligence or AI-enabled digitalization), meaning data-driven, AI-integrated industrial transformation. High-level forums discussed agent guardrails, safety by design, and cybersecurity, but industry sessions repeatedly returned to high-quality datasets, sharing infrastructure, and agent-facing data management as bottlenecks to adoption.

I mentioned in an earlier paper that China’s data policy was centered on three moves: data assetization, data exchanges, and public data franchising. The conferences highlighted several new developments that are woven into China’s data governance strategy:

• Scenario-based data management. Speakers repeatedly argued against building datasets before identifying potential use cases. Their preferred logic is to start from concrete demand—an industrial, administrative, or clinical scenario—and work backward to compile data catalogs, sharing mechanisms, and model architecture. Palantir’s Forward Deployed Engineer position and its ontological modelling came up repeatedly as instruments to specify use scenarios and manage corporate data resources.

• High-quality, interoperable datasets as the bottleneck. The industrial breakthroughs will depend on access to private-domain data and the ability to make that data usable across organizations. The issue is not simply quantity, but interoperability: ownership, structure, format, annotation, and domain context all vary. Several speakers described traits of future data systems designed for AI agents rather than conventional databases: multimodal data, model-data resonance, and ontological modeling.

• Trusted data spaces. This concept, which is being widely tested across sectors, refers to a secure environment where participants share datasets on premises to build products or models, then export the resulting products without seeing the original datasets. I visited one in Shanghai; it resembles a normal computer room with additional security settings. The biggest obstacle is incentives: companies are reluctant to place their key data in the space. Some speakers, e.g., Gao Xinmin (高新民), emphasized adding a data interoperability layer and building industry-specific data commons to incentivize sharing.

2. Health AI: sustainable innovation depends on commercialization and data interoperability

Health is a priority for both China’s data and AI strategies because of its public importance and transformative potential. Better data sharing could improve diagnosis, reduce errors, streamline triage, expand access in grassroots hospitals, and support continuous patient management. The sector also appears more willing to cooperate than many others. But implementation bottlenecks are clear, especially the lack of a payment model for AI-health products. Across two health forums and conversations with five practitioners in hospitals, medical-device companies, and health-data firms, a shared business logic emerged:

At this stage, AI saves money but does not grow the pie for hospitals and pharmaceutical companies. AI’s revenue potential derives from cost savings generated by greater efficiency and accuracy and by reducing trial and error in diagnostics, operations, and research.

This observation has important implications for the business strategies of AI-health enterprises:

• The key question is: Who pays for innovation? Hospitals may receive payment for releasing datasets, primarily from technology and pharmaceutical companies. Technology companies need medical data for model training, while pharmaceutical companies need it to collect post-market real-world evidence and satisfy regulatory requirements. For both groups, hospitals may be the entry point for scaling products, but the real payer is often public health insurance (医保), with commercial insurance playing a secondary but expanding role.

• The industry should no longer expect 医保 to reimburse AI products in a separate category. To commercialize, companies must instead prove that deploying their AI products saves the insurance system money overall—what insiders call “big-picture accounting,” or health economics (算大账). To do that, they must use AI to avoid what already costs the system real money: missed or mistaken diagnoses, unnecessary procedures, etc.

• AI substitution has to be attached to a reimbursable object or process rather than replacing the doctor. Replacing doctors, or increasing efficiency so that they can see more patients, does not by itself create a sufficient revenue stream for AI-health companies. Revenue can be substantial, however, when AI replaces or is bundled into something already paid for—auxiliary devices, consumables, diagnostic equipment, or inefficient procedures. This is why specialized, single-purpose AI is often more monetizable than general medical foundation models, and why many industry-facing medical AI companies are more interested in the former. One example I heard was imaging AI sold to Philips: once embedded as part of the hardware cost, it can be reimbursed through the medical-device pathway.

• The AI + health business model cannot simply be to “build a general medical model and wait for adoption.” It must identify where AI can substitute for costly processes, prove savings at the insurance-system level, and then secure reimbursement so that adoption can scale.

Beijing Hardens Language, Operations at Second Thomas Shoal

On July 20, Chinese and Philippine military personnel came to blows alongside the BRP Sierra Madre, the vessel the Philippines grounded at Second Thomas Shoal (仁爱礁, which the Philippines calls Ayungin Shoal) in 1999 in the South China Sea (SCS). The Philippine military said China Coast Guard (CCG) personnel in a rigid-hull inflatable boat struck a Philippine Navy sailor on the head with a wooden baton, injuring the sailor and damaging the boat. The CCG said the two Philippine boats ignored warnings and rammed its patrol craft, and that “Philippine personnel first used oars, long poles, and other implements to maliciously attack Chinese law enforcement personnel.”

Neither side has released full unedited footage, and the Philippine military says the version Beijing circulated was “spliced or incomplete.”

China Coast Guard Personnel Strike Philippine Navy With Wooden Batons in Ayungin Shoal Incident

The clash broke two years of relative calm. Under the provisional resupply arrangement reached in July 2024, the Philippines ran 13 rotation-and-resupply missions through March 4, 2026—roughly every six weeks—and Philippine officials said those runs passed without dangerous encounters.

The clash also overlaps with the 10th anniversary of the 2016 UNCLOS arbitral tribunal award, which struck down China’s expansion claims in the SCS, including over Second Thomas Shoal. On July 11, the United States and 13 other governments issued a joint statement to commemorate the award, calling on China to respect the findings. China’s foreign ministry responded the next day, calling the award “an illegal and void scrap of paper with no binding force” (一张非法无效没有拘束力的废纸). Chinese state media went further, calling the arbitration a “political farce dressed up in legal disguise.”

Among other things, the 2016 arbitral tribunal ruled that:

• Second Thomas Shoal is a low-tide elevation, meaning it cannot be claimed as sovereign territory by any state on its own.

• The shoal lies within the Philippines’ exclusive economic zone.

• China’s “nine-dash line” claim has no legal basis under UNCLOS to the extent it exceeds the convention’s maritime entitlements.

• China interfered with the Philippines’ lawful exercise of its rights in the SCS, including on Second Thomas Shoal.

China continues to claim the area as part of its “historic rights,” and has maintained that Philippine activities there are unlawful. As a result, there have been repeated confrontations between Chinese and Philippine vessels around the shoal.

CCA Analysis: Beijing is hardening its rhetoric and response to the Philippines’ exercise of its lawful maritime rights on Second Thomas Shoal. The CCG’s July 20 statement included new language not found in previous announcements regarding Second Thomas Shoal, labeling the incident as a “malicious attack” (恶意攻击). The PRC foreign ministry called Philippine actions a “violent assault” (暴力袭击). For comparison, in June 2024, when Chinese personnel boarded Philippine boats armed with knives and axes and severed a Filipino sailor’s thumb, CCG statements only referred to a “dangerous approach” and “deliberate collision,” and the PRC defense ministry acknowledged only that stones had been thrown.

The wording matters because of the claim riding on it. On July 21, the CCG said it had “permitted” the Philippines to move the wounded sailor by small boat to a Philippine coast guard vessel, and that it had “supervised the transfer from start to finish.” Manila rejected the framing a day later, saying the Philippines “does not need, nor does it require, any foreign nation’s permission” inside its own exclusive economic zone, and any claim of permission is “a false narrative intended to project illegitimate authority.”

The PLA appears to have stayed out of the fray. The operation was carried out by the CCG and framed by the foreign ministry, a familiar division of labor that sustains pressure while staying below the threshold of kinetic conflict that might invoke the U.S.-Philippines Mutual Defense Treaty. China’s defense ministry did not issue a statement and did not respond to press queries about the incident. The PLA Southern Theater Command issued a statement after the clash, protesting Philippine-organized “joint patrols” with “outside countries” but did not mention Second Thomas Shoal.

Beijing’s rejection of the arbitration ruling remains unchanged. A decade after the landmark binding judgement, China’s intransigence has caused material harm to its relations with the Philippines in diplomatic and soft power terms. But it has allowed the CCG and PLA to hone a new brand of gray-zone coercion and operational tactics. From Beijing’s perspective, the benefits of rejecting the ruling outweigh the diplomatic blowback, because acceptance would deprive China of its core claims to waters and territory in the SCS.

PLA Initiates Live-Fire Artillery Drills near Taiwan

From July 23 to 24, the PLA Eastern Theater Command (ETC) conducted live-fire artillery drills near Taiwan, in apparent response to a speech by Taiwanese President Lai Ching-te. The exercise occurred east and south of Dongshan Island, off Fujian province, and involved troops from the ETC’s 73^rd Group Army. It reportedly involved the amphibious assault ship Hubei, which undertook landing and take-off training and employed army helicopters.

The drills appeared to be in response to a speech given by Lai during the DPP’s annual party convention several days earlier, in which Lai called for Taiwan to oppose a “red terror” by China, adding that the DPP would never allow a “democratic Taiwan” to “regress into a ‘China’ Taiwan.” The speech drew criticism from the PRC’s Taiwan Affairs Office, which said the remarks “expose Lai’s unchanging ‘Taiwan independence’ stance,” and that Lai had “recklessly challenged the ‘One China’ principle.”

CCA Analysis: China has boxed itself in by responding militarily to every major speech by Lai. Digging in its heels may stoke nationalist sentiment in China and help justify increased PLA training requirements near Taiwan. But this approach also takes cross-Strait relations down a narrower path with fewer avenues for de-escalation that do not involve military escalation. 2028 remains a pivotal year in both Taiwan and the United States, with presidential elections determining the future trajectory of peace and conflict across the Strait. Yet there is little evidence that Beijing is amenable to creative approaches or policies that involve engagement with Taiwanese authorities.

China’s Third Hospital Ship Sails on Its First Overseas Mission

On July 22, the PLA Navy hospital ship Auspicious Ark (吉祥方舟, hull number 868) departed a naval port in Qingdao for Africa and South Asia to carry out Mission Harmony-2026 (和谐使命-2026), the vessel’s first overseas medical deployment.

This is the twelfth Mission Harmony deployment since 2010. The first ten were conducted by the Peace Ark (和平方舟), while the Silk Road Ark (丝路方舟) conducted the eleventh, a 234-day voyage covering roughly 36,000 nautical miles across the Pacific Islands and Latin America—the longest in the series to date. The Auspicious Ark is the PLA Navy’s third 10,000-ton-class hospital ship, first unveiled in May 2025. With its entry into service, each of the PLA Navy’s three theater commands now operates one hospital ship.

CCA Analysis: Humanitarian deployments are not in themselves a new development. It is the tempo that merits attention. Deployments of three different vessels in two years represent a high operational tempo. The Peace Ark returned to port in January 2025; the Silk Road Ark deployed from September 2025 to April 2026; and the Auspicious Ark’s current deployment is scheduled to run into early 2027. With three hulls available, Beijing can now keep one ship forward-deployed on a near-continuous basis while a second undergoes refit and a third remains on domestic duty. What was once an episodic goodwill voyage has become a routine, standing presence.

The tonnage and configuration of these vessels are calibrated for long-term deployments. According to the Center for International Maritime Security, the U.S. Navy’s USNS Comfort and USNS Mercy each carry roughly 1,000 beds and 12 operating rooms, compared to approximately 300 beds and eight operating rooms aboard the Peace Ark. This is an instrument for securing strategic access rather than generating wartime medical capacity, providing a relatively low-cost means of building familiarity with foreign ports and cultivating local goodwill.

The program’s implementation details reveal its broader significance. During its last deployment, the Silk Road Ark conducted joint drills with the Fijian, Tongan, and Brazilian navies in addition to providing outpatient medical work. The PRC Ministry of National Defense has described the Auspicious Ark as having “a functional layout adapted to the demands of preparing for combat” (具有适应备战打仗需求的功能布局). Should Mission Harmony-2026 follow the same template, it would amount to naval engagement with doctors aboard and would represent an opportunity for the PLA Navy to rehearse wartime medical evacuation and treatment through military diplomacy.

Competition is inevitable—and, when governed wisely, can be a powerful driver of innovation, resilience, and human progress. Strategic instability, however, is neither inevitable nor desirable. It can and should be managed through better institutions, clearer guardrails, and sustained dialogue. The central challenge of statecraft in the age of artificial intelligence is therefore not whether to compete, but how to govern competition wisely.

Technology competition between the United States and China has entered a fundamentally new phase. Artificial intelligence, semiconductors, biotechnology, critical minerals, cloud infrastructure, and advanced manufacturing no longer operate as separate policy domains. They increasingly form an integrated strategic ecosystem in which decisions affecting one technology shape outcomes across many others. Yet the institutions governing this competition remain fragmented across agencies, regulatory authorities, and policy silos that were designed for an earlier era.

Three structural shifts have outpaced the existing governance architecture.

First, technology competition has become an integrated governance challenge. Export controls, industrial policy, AI governance, biosecurity, critical minerals, and supply-chain resilience are increasingly interconnected. Policies designed in isolation now generate cascading effects across the broader technology ecosystem.

Second, artificial intelligence has fundamentally changed the nature of technology governance. AI is no longer simply another strategic technology. As a general-purpose capability, it accelerates innovation across semiconductors, biotechnology, scientific research, defense, and advanced manufacturing while introducing new risks related to frontier models, cybersecurity, biosecurity, and CBRN misuse. These challenges cannot be addressed through export controls alone. They require integrated governance.

Third, bureaucratic fragmentation has become a strategic vulnerability. On both the U.S. and Chinese sides, responsibility for technology policy is distributed across multiple institutions with different authorities, incentives, and technical expertise. As technologies converge, fragmented governance limits strategic coordination, slows policy adaptation, and increases the risk of reactive escalation.

These structural changes require a new institutional response.

This special issue proposes establishing a Board of Technology Stability (BoTS) as a permanent mechanism for governing technology competition across export controls, AI governance, biosecurity, semiconductor policy, critical minerals, and supply-chain resilience. Rather than creating another layer of bureaucracy, the Board would integrate existing capabilities, align political leadership with technical expertise, and provide a standing framework for managing strategic competition with greater discipline, coherence, and strategic foresight.

The Board should pursue four immediate priorities.

First, modernize export controls through a standing technical review process that keeps pace with technological change while preserving restrictions on genuinely strategic capabilities.

Second, strengthen supply-chain stability through reciprocal guardrails for critical minerals and rare earths, including civilian end-use assurances, trusted-company licensing arrangements, and longer planning horizons that reduce unnecessary uncertainty while protecting national-security interests.

Third, place AI governance and biosecurity at the center of technology strategy by integrating frontier-model evaluation, AI safety, biosecurity, cybersecurity, and CBRN risk management into a common governance framework.

Fourth, establish a permanent technology dialogue to support crisis communication, improve policy transparency, and develop confidence-building measures where mutual interests exist, particularly regarding AI safety, biosecurity, and strategic supply chains.

The anticipated meeting between Presidents Trump and Xi provides a timely opportunity to launch such an initiative. A Board of Technology Stability would complement emerging trade and investment mechanisms by providing a durable institutional architecture for the defining technologies of the twenty-first century. Its purpose is not to reduce competition, but to ensure that competition strengthens innovation rather than undermines stability.

Competition is inevitable—and welcome. Strategic instability is optional. Through stronger institutions, wiser governance, and sustained dialogue, technology competition can become not only a source of national strength, but also a foundation for long-term strategic stability.

I. Introduction

Technology competition is rapidly becoming the defining feature of the U.S.–China relationship. Increasingly, it shapes not only economic competitiveness but also national security, industrial policy, scientific leadership, and the future balance of geopolitical influence. Yet while technology itself has evolved dramatically, the institutions governing it have changed far more slowly.

Artificial intelligence, semiconductors, biotechnology, cloud infrastructure, critical minerals, advanced manufacturing, and supply chains have become deeply interconnected. Export controls on semiconductor manufacturing equipment now influence AI capability; AI accelerates advances in biotechnology; critical minerals underpin both advanced computing and clean technologies; and supply-chain resilience has become inseparable from economic security and national security. Decisions that once affected individual sectors now increasingly reverberate across the broader technology ecosystem.

Existing governance structures were not designed for this rapidly evolving environment.

Most technology policies continue to be developed through separate bureaucratic channels responsible for export controls, industrial policy, AI regulation, public health, scientific research, investment, diplomacy, and national security. These institutions remain highly capable within their respective domains, but increasingly lack mechanisms for integrated decision-making across technologies whose strategic implications have become inseparable.

The rapid emergence of frontier AI further amplifies this challenge. Artificial intelligence is no longer simply another strategic technology. It is becoming a general-purpose capability that accelerates scientific discovery across nearly every frontier discipline, including biotechnology, materials science, robotics, and defense. At the same time, AI introduces new governance challenges involving frontier-model evaluation, cyber resilience, AI-enabled biological design, and CBRN misuse. These issues cannot be addressed through export controls alone. They require sustained coordination among technical experts, regulators, security agencies, and diplomatic institutions.

Institutional fragmentation compounds these challenges. On both sides, responsibility for technology policy is distributed across multiple ministries and agencies with overlapping responsibilities but limited mechanisms for integrated strategic coordination. As technologies become increasingly interconnected, fragmentation itself becomes a source of strategic risk.

The consequences are already visible. Export controls have been followed by reciprocal restrictions on critical minerals and rare earths. Licensing decisions increasingly shape investment, industrial planning, AI infrastructure, and global supply chains. Individual measures may each be justified on national-security grounds, yet collectively they reveal a broader reality: technology competition has become an integrated governance challenge rather than a collection of discrete policy disputes.

The central question facing policymakers is therefore no longer whether strategic competition should continue. It almost certainly will. The more important question is whether existing institutions remain capable of governing that competition in ways that strengthen national security while preserving innovation, reducing unintended economic disruption, and maintaining strategic stability.

This special issue argues that they do not. Instead, it proposes establishing a Board of Technology Stability as a permanent institutional mechanism capable of integrating technology governance across export controls, AI governance, biosecurity, critical minerals, supply-chain resilience, and strategic dialogue. Rather than replacing existing agencies, the Board would bridge political authority and technical expertise, align policies that are currently managed through separate bureaucratic channels, and provide a durable framework for governing technology competition in the age of artificial intelligence.

II. Three Governance Gaps in the AI Era

The United States does not lack technology strategy. Nor does it lack technical expertise. What it increasingly lacks is an institutional framework capable of governing technology competition as an integrated strategic challenge.

Three governance gaps have emerged as technology competition has evolved. Together, they explain why existing institutions, while individually effective, are becoming collectively inadequate.

The Integration Gap

Technology competition is no longer organized around individual industries or technologies. Artificial intelligence, semiconductors, critical minerals, cloud infrastructure, biotechnology, advanced manufacturing, and supply chains increasingly reinforce one another.

Semiconductor manufacturing capacity shapes AI development. AI accelerates biotechnology and scientific discovery. Critical minerals underpin advanced computing, clean technologies, and defense manufacturing. Cloud infrastructure determines access to computing power. Supply-chain resilience influences both national security and economic competitiveness.

These interactions have transformed technology competition into an interconnected strategic ecosystem.

Policy, however, remains organized around individual sectors and agencies. Export controls, industrial policy, AI governance, biosecurity, investment screening, and diplomacy continue to operate largely in parallel. Decisions taken within one policy domain increasingly generate consequences across many others.

The result is an expanding integration gap between the technologies themselves and the institutions responsible for governing them.

The Capability Gap

The second gap reflects the emergence of artificial intelligence as a general-purpose technology.

Unlike previous generations of strategic technologies, frontier AI accelerates progress across virtually every scientific and industrial domain. It influences semiconductor design, pharmaceutical discovery, biotechnology, autonomous systems, cyber operations, advanced manufacturing, and military capability simultaneously.

This convergence fundamentally changes the nature of technology governance.

National security now depends not only on controlling access to hardware but also on understanding frontier-model capabilities, evaluating AI systems, mitigating cyber vulnerabilities, managing AI-enabled biological risks, and establishing safeguards against CBRN misuse.

These challenges extend well beyond traditional export-control institutions.

No single agency currently possesses the authority or expertise to integrate AI governance, model evaluation, biosecurity, export controls, scientific research, industrial policy, and international engagement into a coherent strategic framework.

As AI continues to evolve, this capability gap will become increasingly consequential.

The Coordination Gap

The third challenge is institutional.

On both sides, technology governance remains fragmented across multiple ministries and agencies.

In the United States, Commerce, Treasury, State, Defense, Energy, HHS, NIST, the White House, and the Intelligence Community each contribute essential capabilities but operate under different statutory authorities, organizational cultures, and policy objectives.

China faces a comparable challenge. Technology policy spans MOFCOM, NDRC, MIIT, CAC, MOST, the National Health Commission, the Ministry of Public Security, the Ministry of State Security, and the broader Party leadership.

These institutions were designed to optimize individual policy domains rather than govern an increasingly interconnected technology ecosystem.

The consequence is an expertise-authority gap.

Technical institutions often possess the deepest understanding of frontier technologies but lack the authority to coordinate government-wide policy. Political leaders possess decision-making authority but rarely receive integrated assessments spanning AI, semiconductors, biotechnology, critical minerals, industrial policy, and supply-chain resilience simultaneously.

Technology competition therefore risks becoming increasingly reactive rather than strategic.

Why the Current Framework Is Under Strain

These three governance gaps increasingly reinforce one another.

The October 2022 semiconductor controls illustrate both the strengths and limitations of the existing framework. They successfully protected the most advanced technologies while demonstrating unprecedented coordination among allies. At the same time, rapid technological progress has outpaced portions of the original framework. Chinese manufacturers have continued advancing beyond several original thresholds, while subsequent licensing decisions—including H200-class GPUs, high-bandwidth memory controls, and repeated “is informed” letters—have required increasingly case-specific adjustments.

At the same time, China’s response has expanded beyond semiconductors. Export controls on rare earths and magnets have demonstrated that supply-chain leverage now operates in both directions. Decisions affecting semiconductor manufacturing increasingly shape critical-mineral policy, investment decisions, AI infrastructure, and industrial planning.

These developments do not suggest that export controls have failed. Rather, they demonstrate that technology competition has become too interconnected to be governed through export controls alone.

The central policy challenge has therefore shifted. It is no longer simply how to design better export controls. It is how to govern strategic competition across an increasingly integrated technology ecosystem while preserving national security, sustaining innovation, reducing unintended economic costs, and maintaining strategic stability.

This is the role a Board of Technology Stability is designed to perform.

III. A Board of Technology Stability

The three governance gaps identified above point to a common conclusion. The United States does not need another export-control mechanism. It needs a governance architecture capable of integrating technology policy across the strategic technologies that increasingly define economic competitiveness, national security, and geopolitical influence.

This paper therefore proposes establishing a Board of Technology Stability (BoTS) with the Chinese side as a permanent institutional mechanism for governing technology competition in the age of artificial intelligence.

The objective of the Board is straightforward: to strengthen national security while sustaining innovation, reducing unintended economic disruption, and improving strategic stability.

BoTS would not replace existing departments or regulatory authorities. Nor would it weaken legitimate competition or dilute national-security protections. Instead, it would provide the institutional capacity to coordinate policies that are currently managed through fragmented bureaucratic structures despite becoming increasingly interconnected.

Technology competition has become systemic. Technology governance must do the same.

Managing Competition, Not Technology

The phrase Technology Stability deserves clarification.

The Board is not intended to stabilize technological progress. Innovation should continue to accelerate.

Nor is it intended to stabilize geopolitical competition. Strategic competition between major powers is likely to remain a defining feature of international affairs.

Rather, Technology Stability refers to reducing unnecessary instability generated by fragmented governance.

Its purpose is to make competition more disciplined, more predictable, and better aligned with long-term strategic objectives.

Competition without governance increases strategic risk. Governance without competition weakens innovation.

The Board seeks to reconcile both.

Why a New Institution?

Governments have repeatedly created new institutions when technological change has fundamentally altered the strategic landscape.

The National Security Council integrated military, diplomatic, and intelligence decision-making during the early Cold War.

The Financial Stability Board emerged after the global financial crisis to strengthen coordination across national regulators.

The proposed Board of Technology Stability reflects a similar need.

Artificial intelligence, semiconductors, biotechnology, cloud infrastructure, critical minerals, and supply chains now interact too closely to be governed independently.

Existing institutions remain indispensable.

But none is responsible for integrating the technology ecosystem as a whole.

The Board would fill that gap.

Five Core Functions

The Board should perform five permanent functions.

First, Integrated Strategic Assessment

Every major technology decision should begin with an integrated understanding of its implications across the broader technology ecosystem.

Rather than assessing export controls, AI governance, critical minerals, biotechnology, and supply chains separately, the Board would evaluate their interaction and long-term strategic consequences.

Integrated assessment should become the foundation of technology statecraft.

Second, Adaptive Export Controls

Export controls remain an indispensable national-security instrument.

Their effectiveness, however, depends upon remaining targeted, technologically current, and strategically coherent.

The Board would establish a standing review mechanism to reassess technological thresholds, evaluate commercial consequences, identify second-order effects, and ensure that controls continue protecting genuinely strategic capabilities while minimizing unnecessary costs to U.S. innovation and allied supply chains.

Export controls should evolve as technology evolves.

Third, AI Governance and Biosecurity

Artificial intelligence has become both the most powerful innovation platform and the most significant governance challenge of the coming decade.

The Board would integrate frontier-model evaluation, AI safety, biosecurity, cybersecurity, CBRN safeguards, and scientific risk assessment into one coherent policy framework.

This would close one of today’s largest institutional gaps: the separation between technical AI expertise and national-security decision-making.

Rather than treating AI governance as a parallel policy discussion, it should become a central pillar of technology strategy.

Fourth, Supply-Chain Stability

Technology competition increasingly operates through global production networks rather than individual firms.

Critical minerals, semiconductor manufacturing equipment, advanced manufacturing, cloud infrastructure, and logistics have become sources of both resilience and strategic leverage.

The Board would coordinate policies affecting these interdependent systems while reducing unnecessary uncertainty for governments, businesses, and allied partners.

Supply-chain resilience should become a strategic objective rather than a crisis response.

Fifth, Strategic Dialogue

Finally, the Board should establish a permanent technology dialogue between the United States and China.

Strategic competition requires communication as much as deterrence.

Technical experts should have regular channels to discuss emerging technologies, AI safety, biosecurity, supply-chain resilience, and reciprocal concerns before they escalate into political crises.

Dialogue should not be understood as an alternative to competition.

It is one of the mechanisms through which competition can be governed responsibly.

A September Deliverable

The anticipated meeting between Presidents Trump and Xi offers an opportunity to move beyond tactical negotiations toward institutional innovation.

Recent discussions have focused appropriately on tariffs, trade, export controls, investment, and critical minerals.

These issues remain important. But they are increasingly symptoms of a larger structural transformation.

Artificial intelligence is reshaping every dimension of technology competition. Biosecurity has become inseparable from AI governance. Critical minerals have become strategic assets. Supply chains have become instruments of national power.

These developments require institutions capable of governing competition across the technology ecosystem rather than negotiating each issue separately.

A Board of Technology Stability would provide precisely such a framework.

It would complement the proposed Boards of Trade and Investment by focusing specifically on the technologies that increasingly determine economic security, national security, scientific leadership, and strategic stability.

More importantly, it would demonstrate that both governments recognize a fundamental reality:

Technology competition is likely to endure.

Strategic instability does not have to.

IV. Immediate Priorities

The Board of Technology Stability should focus initially on a limited set of priorities where technological change has outpaced existing policy and where coordinated action would strengthen national security, preserve innovation, and reduce strategic instability. Early success will depend less on creating new regulations than on improving coordination across existing institutions.

Priority One: Modernize Export Controls for the AI Era

Export controls remain one of the United States’ most important national-security tools. The October 2022 semiconductor controls demonstrated Washington’s ability to protect the most advanced capabilities while coordinating closely with key allies. Those objectives remain valid.

Technology, however, has evolved rapidly. Chinese manufacturers now operate beyond several of the original memory and logic thresholds, while subsequent adjustments—including licensing decisions for H200-class GPUs, new controls on high-bandwidth memory, and repeated “is informed” letters—illustrate the increasing difficulty of governing frontier technologies through static regulatory frameworks.

The objective should therefore be continuous adaptation rather than periodic revision.

The Board should establish a permanent technical review mechanism that regularly evaluates technological thresholds, commercial consequences, allied coordination, and Chinese responses. Export controls should remain targeted, technologically current, and strategically coherent. Their effectiveness should be measured not only by restricting access to frontier capabilities, but also by preserving U.S. innovation capacity, sustaining allied technological leadership, and minimizing unintended incentives for accelerated localization.

Priority Two: Stabilize Strategic Supply Chains

Supply-chain resilience has become a central dimension of technology competition.

Recent Chinese export controls on rare earths and permanent magnets demonstrate that strategic leverage increasingly extends beyond semiconductors. At the same time, the experience of the past several years has shown that not all technological chokepoints are alike.

Rare earth processing, specialized equipment, and magnet production remain highly concentrated and difficult to replicate. Semiconductor innovation, by contrast, is characterized by globally distributed ecosystems, multiple technology pathways, and strong incentives for substitution. Policy should distinguish between these fundamentally different forms of technological dependence.

The Board should therefore pursue a framework of strategic supply-chain stability.

Possible measures include reciprocal civilian end-use assurances, trusted-company licensing arrangements, multi-year licensing horizons, and structured dialogue regarding critical-mineral supply chains. Such measures would preserve legitimate national-security protections while reducing unnecessary uncertainty for governments, companies, and global markets.

The objective is not to eliminate leverage. It is to reduce unnecessary volatility in sectors critical to both strategic competition and global economic stability.

Priority Three: Place AI Governance and Biosecurity at the Center of Technology Strategy

Artificial intelligence should no longer be treated as a specialized technology issue. It has become the enabling platform for virtually every frontier technology.

Accordingly, AI governance and biosecurity should become central responsibilities of the Board rather than parallel policy discussions.

The Board should establish an integrated framework linking frontier-model evaluation, AI safety, cybersecurity, biosecurity, CBRN safeguards, and scientific risk assessment. Technical expertise from institutions such as CAISI/NIST, DOE, HHS, NIH, CDC, ASPR, BARDA, the national laboratories, and other relevant agencies should be coordinated within a single strategic process capable of informing national-level decision-making.

Internationally, the Board should support sustained technical engagement with relevant counterparts in China, including ministries, standards organizations, AI safety institutions, and public-health authorities. Such engagement would not reduce strategic competition. Rather, it would strengthen confidence-building measures in areas where mutual interests exist, particularly concerning frontier AI safety, biosecurity, and catastrophic-risk prevention.

In the age of AI, governance of biological risk should become a core function of technology strategy rather than an afterthought.

Priority Four: Institutionalize Technology Dialogue

Strategic competition increasingly unfolds through technology policy.

Yet bilateral engagement remains largely reactive, organized around successive disputes over export controls, tariffs, investment restrictions, or critical minerals.

The Board should establish a permanent channel for technology dialogue that complements existing diplomatic and economic mechanisms.

This dialogue would enable both governments to communicate policy intentions, discuss emerging technologies before they become crises, coordinate technical assessments where appropriate, and develop reciprocal guardrails in areas of shared interest.

Dialogue should not be viewed as a concession. Nor should it be confused with cooperation.

It is a mechanism for governing competition responsibly.

As strategic competition intensifies, communication becomes more—not less—important.

Priority Five: Build Institutional Capacity for Continuous Adaptation

Perhaps the Board’s most important function is also the least visible.

Technology evolves far more rapidly than government institutions. Policy reviews conducted every several years are increasingly insufficient for technologies whose capabilities can change within months.

The Board should therefore institutionalize continuous learning.

Standing technical working groups, regular strategic assessments, horizon-scanning exercises, engagement with industry and academia, and structured collaboration with allies should become permanent features of technology governance rather than ad hoc responses to individual crises.

The objective is to create an institution capable not merely of responding to technological change, but of adapting alongside it.

In an era of accelerating innovation, institutional agility is itself a strategic advantage.

V. Institutional Design

A Board of Technology Stability should be designed around one principle:

Integrate expertise without centralizing bureaucracy.

The proposal is not intended to create another large department or replace existing agencies. Rather, it seeks to provide the institutional capacity that existing structures currently lack: the ability to integrate political authority, technical expertise, and strategic decision-making across interconnected technologies.

Accordingly, the Board should distinguish clearly between political leadership and technical leadership.

Political Leadership

Political leadership should remain at the Cabinet level.

Its role is to establish strategic priorities, resolve interagency disagreements, oversee implementation, and ensure that technology policy remains aligned with broader national-security and economic objectives.

Given Treasury’s expanding role in economic security and international negotiations, the Secretary of the Treasury would be well positioned to provide overall political leadership while working closely with the White House, Commerce, State, Defense, and other relevant departments.

On the Chinese side, an equivalent political lead would require sufficient authority to coordinate across industrial policy, economic policy, AI governance, and national security. The key point is institutional: successful governance requires a political leader capable of integrating across bureaucratic boundaries.

Technical Leadership

Political leadership alone is insufficient.

The Board should also be led by a senior technical figure with deep expertise across AI, semiconductors, biotechnology, critical minerals, industrial policy, and global technology ecosystems.

Unlike traditional advisory structures, this role should be permanent, independent of individual departments, and capable of translating technical analysis into strategic policy recommendations.

The central purpose is to close the longstanding gap between expertise and authority.

Functional Organization

Rather than organizing the Board around agencies, it should be organized around five enduring functions:

• Technology Strategy and Integrated Assessment

• Export Controls and Industrial Competitiveness

• AI Governance, Frontier Models, and Biosecurity

• Critical Minerals and Supply-Chain Resilience

• Strategic Dialogue and International Coordination

This functional structure would encourage integrated decision-making while preserving the responsibilities of existing departments.

U.S.–China Institutional Mapping

The Board should also maintain a standing institutional map identifying the principal counterparts on both sides.

On the U.S. side, this includes Commerce, Treasury, State, Defense, Energy, HHS, NIST, the Intelligence Community, and the White House.

On the Chinese side, it includes MOFCOM, NDRC, MIIT, CAC, MOST, the National Health Commission, and the relevant Party leadership, together with technical institutions responsible for AI safety, standards, biosecurity, and industrial policy.

Detailed personnel assignments should remain flexible and evolve alongside technological developments. The Board’s strength lies not in organizational charts, but in its ability to convene the right expertise at the right time.

VI. Conclusion

Governing Competition in the Age of AI

Competition between the United States and China is likely to remain a defining feature of the international system for the foreseeable future. Artificial intelligence, semiconductors, biotechnology, critical minerals, advanced manufacturing, and cloud infrastructure have become central to economic competitiveness, national security, and geopolitical influence. As these technologies converge, technology competition itself is becoming increasingly systemic.

The institutions governing that competition, however, have not evolved at the same pace.

For much of the past two decades, technology policy could be managed through relatively discrete instruments—export controls, industrial policy, investment review, scientific cooperation, and trade negotiations. Today, these instruments increasingly overlap. Decisions affecting semiconductors influence AI capability. AI accelerates biotechnology. Critical minerals shape industrial resilience. Supply chains have become instruments of both economic efficiency and strategic leverage.

Technology competition has become an integrated governance challenge.

This paper has argued that three governance gaps now limit effective policymaking.

The first is an integration gap between increasingly interconnected technologies and institutions that continue to govern them separately.

The second is a capability gap between the challenges created by frontier AI and biosecurity and the institutional tools available to manage them.

The third is a coordination gap between political authority and technical expertise, both within governments and across international partners.

These gaps are unlikely to narrow on their own.

They require institutional innovation.

The proposed Board of Technology Stability is intended as that innovation.

Its purpose is neither to reduce strategic competition nor to slow technological progress. Rather, it seeks to strengthen national security while preserving innovation, improving policy coherence, reducing unintended supply-chain disruption, and lowering the risk of unnecessary escalation.

Technology stability should therefore not be understood as technological stagnation.

It should be understood as strategic stability in an era of accelerating technological change.

The Board would complement—not replace—existing institutions by integrating export controls, AI governance, biosecurity, industrial policy, critical minerals, and supply-chain resilience within a common strategic framework. It would connect political leadership with technical expertise, allowing governments to respond to technological change with greater agility, coherence, and foresight.

The anticipated meeting between Presidents Trump and Xi provides an opportunity to begin building such an architecture.

Much attention has rightly focused on tariffs, export controls, investment, and critical minerals. These issues will remain important. Yet they increasingly represent manifestations of a broader transformation rather than isolated policy disputes.

Artificial intelligence is reshaping every dimension of technology competition.

Biotechnology is becoming increasingly inseparable from AI.

Critical minerals have become strategic assets.

Supply chains have become instruments of national power.

Future competition will increasingly be defined by the interaction among these technologies rather than by any single sector.

The challenge facing policymakers is therefore no longer simply how to compete.

It is how to govern competition.

History suggests that major technological revolutions have consistently required new institutions. Nuclear technology led to new arms-control mechanisms. Globalization produced new trade and financial institutions. The AI era will likewise require governance mechanisms capable of integrating scientific innovation, national security, economic policy, and international stability.

Competition without governance increases strategic risk.

Governance without competition weakens innovation.

The challenge of statecraft in the age of AI is to build institutions capable of governing competition wisely.

A Board of Technology Stability represents one practical step toward that objective.

PLA Navy Conducts Unprecedented Submarine-launched Missile Test into Pacific

On Monday, July 6, the PLA Navy (PLAN) carried out a rare submarine-launched ballistic missile (SLBM) test with a dummy warhead from a nuclear-powered ballistic missile submarine (SSBN). The missile was reportedly launched from a location south of Guangdong and flew thousands of miles across the Pacific, landing in the South Pacific near Tonga. China described the test as a “routine” military exercise.

The test is significant because it is the first publicly acknowledged SLBM test with a dummy warhead from the PLAN SSBN that traveled so far into the Pacific (approximately 7,200 kilometers). Previous SLBM tests landed in the Bohai and were not publicly announced.

The test prompted swift condemnation from the U.S., Japan, Australia, New Zealand, and Taiwan, who accused Beijing of “destabilizing” the region.

It appears the missile tested was the JL-2 (巨浪-2), not the JL-3 (巨浪-3). The JL-2 has a maximum range of 7,200–8,000 kilometers, while the JL-3 is 10,000-13,000 kilometers. The range of the missile test is consistent with that of the JL-2.

With this test, the PLAN has shown the capability to target the continental United States from bastions close to Chinese waters, in addition to holding U.S. allies and partners in the region at risk. The launch indicates that China now has an operational sea-based nuclear deterrent and is progressing with its nuclear triad.

CCA Analysis: A test of this range is a major development that indicates that China is moving toward a more survivable and longer-range sea-based nuclear deterrent capability. The launch demonstrates that China’s nuclear deterrent is no longer centered solely on land-based missiles.

This will make Xi Jinping very happy. Conducting a launch of this magnitude implies confidence in China’s submarine crews, command and control, and missile reliability in the context of ongoing political pressure with Xi’s anti-corruption campaign and PLA purges.

It is noteworthy that China pre-notified the United States, Japan, New Zealand, Australia, Papua New Guinea and Taiwan - but reportedly only a few hours in advance and with limited details. The U.S. criticized the short advance notice as “insufficient” and below the notification standards generally maintained by the P5 (meaning the five permanent members of the United Nations Security Council). China publicly stated that it had notified “relevant countries” in advance but did not specify which countries or how much warning they received.

This contrasts with China’s September 25, 2024 ICBM test into the Pacific, when Beijing reportedly notified the United States, Australia, New Zealand, and Japan approximately 24 hours in advance.

The rationale for the test is straightforward: in order to boast a true nuclear triad, China has to be able to demonstrate its missiles work as designed, including reaching the expected range and landing in the pre-determined target zone. In this case, it appears China and the PLAN achieved these objectives.

I wouldn’t read too much into the timing. A launch of this type was likely planned months, maybe years, in advance, and is subject to delays based on weather. Even if China intended to time the launch with a historical event, much depends on submarine crew readiness and weather, which can change at a moment’s notice.

The test will almost certainly facilitate a desire from regional powers, including U.S. allies and partners, to cooperate more, not less, with the U.S., due to greater fear and concern over China’s nuclear modernization trajectory. That is the downside of the launch from Beijing’s perspective. It will bring U.S. allies closer together and fuel concern over China’s military ambitions.

The U.S. State Department’s condemnation of the launch won’t change Beijing’s calculus on arms control or nuclear risk reduction with the United States, or Washington’s efforts to compel Beijing to agree to trilateral arms control treaties with Moscow and Washington. China believes it is still at a strategic disadvantage with the U.S. when it comes to nuclear capabilities, and in that context, will not be persuaded in the short term to agree to any sort of arms control with the U.S.

China and Russia Kick Off Joint Naval Exercises in Qingdao

On July 6, the Chinese and Russian navies launched “Joint Sea-2026” (海上联合-2026), an annual bilateral naval exercise, in Qingdao, China. The exercise is noteworthy because it involves surface combatants, submarines, rescue vessels, and joint air-defense and anti-missile training, and will be followed by a joint China-Russia maritime patrol in the Pacific Ocean.

This year’s exercise consists of an initial harbor phase, followed by at-sea operations beginning July 9. According to Chinese official reporting, the sea phase is taking place in waters and airspace near Qingdao and includes four publicly identified major training areas: joint reconnaissance, air and missile defense, attacks against maritime targets, and joint submarine rescue operations.

The harbor phase includes command and tactical-coordination activities, professional exchanges on submarine rescue technology and equipment, reciprocal ship visits, and other exchanges. Chinese personnel toured Russian ships, while Russian sailors visited the Chinese destroyer Kaifeng.

Russia has deployed a substantial contingent from its Pacific Fleet, including the Slava-class guided-missile cruiser Varyag, a corvette, a diesel-electric submarine, and the submarine-rescue vessel Igor Belousov. The participation of a Russian submarine and specialized rescue vessel is especially relevant given that joint submarine rescue is one of the exercise’s principal training areas.

On the Chinese side, the publicly identified combatants include the PLAN’s Type 052D guided-missile destroyer Kaifeng; the Type 055 guided-missile destroyer Anshan, the Type 054A guided-missile frigate Wuhu; the Type 903A comprehensive replenishment ship Kekexilihu; the Type 926 submarine support ship Yangchenghu, and one submarine that was not identified.

The publicly announced training program is significant because it goes beyond simple formation sailing or basic communications drills. The principal activities are:

• Joint reconnaissance, which includes sharing information and coordinating detection and tracking of simulated threat actors;

• Air and missile defense, which is arguably one of the most strategically significant components, requiring integration of sensors, command systems, and defensive weapons across both navies;

• Maritime strike operations, which include coordinated attacks against surface targets; and

• Joint submarine rescue, which is an area of growing China-Russia technical cooperation and a potentially important indicator of deeper submarine-force interoperability.

During the harbor phase, Chinese and Russian experts specifically discussed submarine rescue technology, equipment development, and maritime training, suggesting that this element is more than a ceremonial addition to the exercise.

Following Joint Sea-2026, some participating Chinese and Russian forces will reportedly proceed into the Pacific Ocean for a joint maritime patrol. The exact patrol route has not yet been publicly specified in the sources available, but previous China-Russia joint patrols have taken their ships through strategically sensitive areas of the western Pacific, including waters near Japan, which at times have unnerved the Japanese military.

This follow-on patrol is as strategically consequential as the exercise itself. A scripted exercise demonstrates interoperability under controlled conditions. But an operational patrol requires the two navies to coordinate over greater distances and periods of time in real-world maritime environments. On July 11, the Global Times posted a video on X of Chinese and Russian submarines conducting at-sea patrols, suggesting a new level of joint undersea warfare training in a live setting.

CCA Analysis: There are four key takeaways from this year’s “Joint Sea-2026” exercise.

First, the exercise demonstrates the continuing institutionalization of China-Russia military cooperation. The two countries have become each other’s most important foreign military exercise partners, and their exercises have progressively expanded in frequency, geographical reach, complexity, and service participation since the mid-2000s.

Second, the air and missile defense component has increased in emphasis. Effective integration of air and missile defense requires coordination in detection, tracking, information sharing, and command and control. The critical analytical question is how much tactical data the two navies actually exchange during the drill. Public information does not suggest that their combat systems are fully integrated. But the scale and scope of bilateral naval exercises is clearly intended to improve their ability to operate alongside one another, and one assumes integration has improved.

Third, Qingdao is home to major elements of China’s Northern Theater Command Navy and is an important PLAN submarine base. Conducting the exercise there and explicitly emphasizing submarine rescue gives Russia access to a strategically important Chinese naval environment that provides both navies an opportunity to deepen cooperation in an area that has traditionally involved particularly sensitive military technologies.

Fourth, the exercise occurs amid a wider surge in Chinese military activity around Taiwan. Taiwan’s National Security Bureau reported an upward trend in Chinese naval movements and said that four Chinese naval formations were operating in the western Pacific as of last week. Taiwanese officials have interpreted expanding China-Russia activity as part of a broader effort to counter the U.S.-led military posture along the First Island Chain.

Thus, the most important aspect of “Joint Sea-2026” is not so much the number of ships involved, but the increasing sophistication and routinization of China-Russia naval cooperation at the tactical level. The inclusion of submarine rescue operations, air and missile defense, maritime strike, reconnaissance, and subsequent operational patrols represents a substantially more meaningful form of cooperation than ceremonial port visits.

At the same time, I would not characterize China and Russia as a fully integrated military alliance. There is still limited public evidence of the kind of deeply integrated command-and-control architecture or joint operational planning associated with the United States and its allies, for example. A better characterization is that China and Russia are developing an increasingly capable, institutionalized, and strategically coordinated military partnership that gives both countries greater capacity to complicate U.S. and allied planning across multiple theaters.

Xi Promotes Two Generals, Including One to CMC; Purges Others

On July 3, Chinese President Xi Jinping promoted two PLA officers to the rank of General, and named a new head of the Central Military Commission (CMC)’s anti-corruption body.

Zhang Shuguang (张曙光), was promoted from the rank of lieutenant general to full general, was named as the new head of the CMC’s Discipline Inspection Commission – the top military anti-corruption body – and director of the CMC’s supervision committee, replacing Zhang Shengmin. Zhang Shuguang spent most of his career as a political officer in the Air Force and Army until working in the DIC under Zhang Shengmin. His elevation to replace Zhang Shengmin is therefore not a surprise, given the patronage relationship that existed prior to his promotion.

Wang Gang (王刚) was also promoted to the rank of full general and takes over as Commander of the PLA Air Force (PLAAF). Wang began his career as a pilot before leading an Air Force division. Wang later served as the director of the Air Force training department (2012–2016), assistant Air Force chief of staff (2016–2019), central theater command Air Force chief of staff (2019–2022), Air Force chief of staff (2022–2025), and deputy Air Force commander (2025–2026).

The announcement comes amid an ongoing anti-corruption campaign that has led to the downfall of many of the most senior figures in the PLA.

PLA analyst Zi Yang covers the backgrounds and significance of these two generals in detail in this piece in The Diplomat.

In other recent developments, on June 26, the National People’s Congress (NPC) Standing Committee announced the removal of six PLA general officers.

• Xu Xueqiang (许学强) (PLAAF), head of the Equipment Development Department and director of the Manned Space Program of PRC;

• Wang Kangping (王抗平) (PLAAF), deputy commander of the PLA Eastern Theater Command;

• Li Fengbiao (李凤彪) (PLAA), political commissar of the Western Theater Command;

• Yin Hongxing (尹红星), (PLAA), political commissar of the Southern Theater Command Army;

• Guo Puxiao (郭普校), (PLAAF), political commissar of the PLA Air Force; and

• Zhang Minghua (张明华), (PLAA), commander of the PLA Cyberspace Force.

What is most noteworthy about these purges is they appear to target the Political Work department of the PLA Air Force (PLAAF) more than in the past. Three are Air Force officers, with three from the PLA Army (PLAA). Up until this point, the PLAAF had largely escaped the full brunt of Xi’s corruption effort, which had disproportionately targeted the PLA Army, Navy, Rocket force, and equipment forces. The Jamestown Foundation’s China Brief has a helpful write-up of these recent purges and their significance.

CCA Analysis: It is safe to say this latest round of purges is focused primarily on dismantling the residual patronage network around the former CMC vice chair Zhang Youxia (张又侠), who became one of Xi’s highest-profile purges of senior PLA officers when he was removed from office on January 24, 2026.

It was also noteworthy that Defense Minister Dong Jun and General Han Shengyan, commander of the Central Theatre Command, were pictured on July 1 seated next to each other at a gala marking the 105th anniversary of the Communist Party’s founding. It was the first time they had been pictured alongside other senior PLA commanders.

This could be an indication that both Dong and Han may be on track to become CMC members as well.

Given the two new promotions to General, if I were to guess, the new CMC leadership in 2027 could include Zhang, Wang, Dong, Han, and Yang Zhibin, who was appointed commander of the Eastern Theatre Command in December. But we will have to wait and see.

Written by Kate Logan, Director of the China Climate Hub

In May, China’s power sector capacity breached 4000 gigawatts (GW) – exceeding the total magnitude of the U.S., EU, Indian, Russian and Japanese power sectors combined. The paradox of this scale, of course, is that China leads the world in clean power installations and is approaching a power sector based primarily on non-fossil sources while still burning the most coal of any country, accounting for over half of the global total.

Subscribe now

Beijing recently released its 15th five-year plan for building a new-type energy system, which introduced little by way of flashy new targets for clean energy or emissions control. But asking why officials avoided any surprises, along with a deeper read of the plan, can reveal useful insights about how domestic policymakers are approaching China’s energy transformation. In particular, three points deserve attention from those aiming to understand how Beijing intends to thread the needle between the triple imperatives of economic growth, energy security, and decarbonization.

First, the plan centers the next phase of China’s efforts to safeguard its energy security as it emerges from the Hormuz energy crisis relatively unscathed.

Recent weeks have seen a flurry of media coverage declaring China to be the “winner” from the Strait of Hormuz energy crisis. While some analysts were surprised by China’s relative ability to weather deep shocks to global energy markets, Beijing emerged from the crisis with greater confidence, reaffirming its conviction to double down on energy independence and self-sufficiency. The new plan places energy security at the center while outlining how officials plan to reduce China’s remaining import dependencies. This focus is evident from the opening paragraphs, which foreground compounding security risks while framing energy security and supply adequacy as the prerequisite for economic viability, technological innovation, and decarbonization.

The document outlines a variety of steps for further substituting imported fuels with domestic alternatives, while pledging that “energy imports will be diversified and controllable.” Major components of this strategy include expanding the development of coal-based synthetic fuels; expanding and commercializing green fuels, including by increasing green hydrogen output eightfold by 2030; and deepening electrification of heavy-duty trucking and marine shipping. In the power sector, natural gas, which already plays a minimal role in electricity generation, is to be constructed “rationally” while scaling up domestically produced gas turbines. China will also aim for 35 percent electrification by 2030, five years ahead of a newly proposed global target.

For China’s emissions, the implications of this strategy are likely mixed. On one hand, further minimizing oil and gas use and accelerating electrification will deepen reductions. However, even with the emphasis on green fuels, renewed political support for coal-to-chemicals investments risks unleashing a massive new emissions “bomb” – and one that may become increasingly difficult to track via official statistics.

Second, the plan’s emphasis on enhancing grid flexibility signals Beijing’s concern over high levels of “wasted” renewables, which are a bottleneck for both economic growth and decarbonization.

With newly installed wind and solar capacity growing at record rates in recent years, China’s relatively inflexible grid has struggled to integrate this sudden influx of intermittent power. Just last year, China added 434 gigawatts of wind and solar, a figure nearly one and a half times the size of Japan’s entire power sector. Official data now place renewable energy curtailment at just under the government’s recently-established limit of 10 percent, which was doubled from 5 percent last year. Analysts, however, believe that the actual curtailment figure is substantially higher than official figures suggest. High curtailment is impeding clean power from keeping up with

China’s rapidly growing power demand, which increased by around 5 percent last year, forcing greater coal consumption than it would be otherwise.

The solution – making China’s power sector more flexible – is at the core of the new plan. It includes a new target for the grid’s “regulating capacity” to rise by over 40 percent, which serves as a proxy for the grid’s adaptability. Details about how this will look in practice are vague, however, and the language seems to indicate Beijing’s intention to continue “bundling” coal alongside cleaner sources of generation. But several additional targets for 2030 provide a strong growth incentive for cleaner means of balancing load variation: raising “new energy storage,” which includes batteries, to 300 GW; increasing pumped-storage hydropower to 160 GW; and expanding vehicle-to-grid capacity and virtual power plant capacity to 50 GW each. Some of these targets fall short of industry association benchmarks, but they are more ambitious than the objectives around clean energy installations – reflecting Beijing’s focus on making the existing system work better before adding even more new energy.

Third, the plan reaffirms Beijing’s shift away from control-based targets and toward an “additive” approach that leverages China’s clean technology dominance as the main driver of decarbonization.

For those hoping for firmer signals on how China will limit fossil fuels and cap its emissions, the new plan will disappoint. Although the overall 15th Five-Year Plan pledges to peak oil and coal consumption – and another recent initiative integrates oil and coal use totals into local cadres’ evaluations – the new energy plan provides no additional details about the timing and level of such peaks. This is despite earlier indications in state media that oil consumption will likely peak around 2026 and coal consumption around 2027. On renewables, the plan mostly consolidates existing targets, including 2030 goals for non-fossil energy to account for 25 percent of total energy consumption and for “new energy generation” – i.e., wind and solar – to reach 30 percent of total power generation. An additional, new target for all non-fossil sources to comprise over half of power generation would still leave room for overall emissions to rise.

From Beijing’s perspective, these conservative renewables targets and the absence of hard caps on fossil fuels allow China the flexibility to grow its economy and weather potential geopolitical shocks while remaining on track to peak its emissions before 2030, as Xi pledged in 2020. They also underscore that Beijing’s level of ambition is likely to be driven less by climate commitments than by economic incentives from its leading clean technology industries, as my colleagues Neil Thomas and Guoguang Wu argued in a recent CCA report. Innovation will be an important part of China’s recipe: the plan puts forth new targets for international patent applications to rise by 5 percent and for over 100 “first-of-its-kind” demonstration projects to be launched by 2030. The plan also elevates a May 2025 policy calling for computing power demand to be met with green sources, marking the first time that “computing power” was referenced in a national energy plan.

Observers may debate whether Beijing’s new energy strategy meets the moment, especially whether it balances economic and security interests with the realities of climate change: in 2025, China suffered over $35 billion in economic losses from natural disasters, and Xi recently chaired a Politburo study session on emergency management in the wake of disastrous flooding in China’s Southern regions this year. Even without ambitious targets, however, Beijing would argue that it has already demonstrated the essential role of electrification for bolstering energy security in a volatile world. With an electrification rate significantly higher than those of the United States or Europe, China may also be better positioned to transform a wide range of industrial sectors while reducing both energy costs and carbon emissions.

Executive Summary

• This analysis argues that the next U.S.-China AI safety dialogue will be constrained not only by mistrust or geopolitical competition, or by lack of political will on both sides, but by a more practical bureaucratic problem: neither government yet has an institutional architecture that combines technical model-evaluation capacity, national-security authority, industrial-policy awareness, diplomatic coordination, and access to top-level political decision-making that could lead to real outcomes enforceable in both countries and acceptable to the global AI safety community.

• The core analytical frame is the expertise-authority gap. In both systems, the actors with the deepest AI model evaluation and testing knowledge often lack regulatory or political authority, while the actors with authority often have not exercised that authority within a broader policy process and critically lack the specialized capacity to evaluate frontier models. This makes even politically endorsed dialogue difficult to translate into a durable process.

• In the United States, CAISI, Commerce/NIST, NSA, CISA, DOE/NNSA, ONCD, OSTP, Treasury, and State all bring some assets to the frontier-AI governance game, but no single institution clearly owns model evaluation, release decisions, let alone bilateral risk-reduction diplomacy. The June 2026 executive order begins to build a classified cyber-focused evaluation process, but it remains voluntary, contested, and incomplete, with multiple players vying for primacy.

• China faces a parallel but differently structured problem. CAC, MIIT, NDRC, MOST, MFA, MSS, MPS, the PLA, and emerging technical networks such as CnAISDA all have relevant equities, yet Beijing still lacks a clearly empowered frontier-AI risk evaluator. This fragmentation matters because effective cooperation will require counterpart capacity: both sides need institutions able to speak the same language, define risk thresholds, evaluate capabilities, and speak with enough authority to sustain agreements.

• Thus what will likely emerge over the what will be at least a one to two year process will not be a single bilateral arms-control treaty or simply a purely technical exchange, but a layered and flexible architecture that could include at least some of the following elements: technical evaluation channels between expert bodies, political oversight through national-security and diplomatic institutions, and narrowly scoped confidence-building measures beginning with risks both sides already recognize, such as cybersecurity, and biosecurity misuse at a minimum, and potentially RSI and loss of control in the most optimistic scenario.

  

Bureaucratic Capacity Matters

This causal claim is deliberately bounded. Trust deficits and strategic competition remain fundamental, but bureaucratic fragmentation explains why even moments of political will may fail to generate operational cooperation, and particular across administrations in the U.S., where there is significant turnover of bureaucratic leadership every four years. Without agencies that can evaluate models, define thresholds, coordinate across economic and security equities, and authorize follow-through, leader-level endorsement produces a channel rather than a process.

The emerging U.S.-China AI governance and safety discussion, one of, if not the most important of the key deliverables of the May 2026 U.S. China Summit in Beijing between Presidents Trump and Xi, has usually been treated as a problem of strategic engagement and trust in the wake of a rapidly developing technology. The new government-to-government talks will commence as Washington and Beijing continue to contest the AI space as governed by bilateral competition over AI hardware, commercial ecosystems, military applications, talent, and the broader geopolitical advantages associated with leadership at the frontier edges of AI. These conditions make any bilateral discussion of AI safety vulnerable to concerns over intent, access, verification, security, and relative advantage. It is telling that the new dialogue was announced as a formal summit outcome by Beijing, while Washington has so far issued no parallel statement of its own, an early indicator of the asymmetries in institutional readiness described below.^[1]

But bureaucratic capacity may be the more immediate constraint through which strategic mistrust becomes operationally binding. As CCA Honorary Fellow Paul Triolo recently wrote, the governance of frontier AI models and the upcoming U.S.-China dialogue will involve interactions between a complex array of agencies, overlapping mandates and authorities, that pose significant institutional and political challenges. Neither the United States nor China has yet built a fully mature institutional system for governing frontier AI as a national-security technology. Both governments have agencies with partial responsibility for the issue, but neither has a settled architecture capable of integrating all the critical dimensions that will be needed: technical evaluation, national-security considerations, industrial-policy impacts, diplomatic coordination, and political decision-making, to name the most obvious. This is a faster-moving version of a familiar institutional lag. The early nuclear age also forced governments to build governance structures only after new technology had transformed strategic competition and there had been a Trinity Moment. But frontier AI is more diffuse and there will not be a Trinity Moment to focus minds on both sides: it is driven by private firms, the first time for a technology of this importance and wide ranging consequences not developed by government or incubated by government deployed through commercial platforms, improved through rapid software iteration, and relevant across cyber, biosecurity, military, economic, and information domains, all at once. Two superpowers are attempting to put guardrails around a rapidly developing technology without the needed bureaucratic scaffolding to both develop and implement agreements. Accelerating the development of this scaffolding both domestic and bilaterally this will be critical for not just the two countries, but the entire sector and the broader world.

To be sure, the two governments have been here once before, and the precedent is instructive. A first intergovernmental dialogue on AI was held in Geneva in May 2024, led on the U.S. side by the NSC’s senior director for technology and national security and the State Department’s then acting special envoy for critical and emerging technology, and on the Chinese side by the MFA’s Department of North American and Oceanian Affairs.^[2] Later that year, Presidents Biden and Xi affirmed in Lima the need to maintain human control over the decision to use nuclear weapons, which remains the only concrete bilateral AI understanding to date.^[3] The Geneva channel and the State Department’s position did not survive the U.S. political transition, primarily because it was staffed by generalist diplomatic and NSC channels rather than by institutions with technical capacity over and authority for frontier models, which, as this piece argues, both sides still lack. Things are somewhat more favorable now for the participation of the requisite technical expertise sorely lacking in Geneva, but there is still no process or clear delegation nor authority that would facilitate real progress. The lesson for the new dialogue then is that leader-level endorsement without bureaucratic scaffolding produces a channel, not a process likely to endure on such a complex topic.

This reality, in the near term, is likely to constitute a major bottleneck for progress on AI safety cooperation. Frontier AI models and platforms do not fit neatly into the categories through which modern states have traditionally managed strategic risk. The most consequential models have capabilities that can spill across traditional national security areas such as cyber operations and critical infrastructure vulnerability, biological agent design, military integration, and intelligence analysis, to broader areas such as scientific discovery, economic growth across most sectors, and future risks such as recursive self-improvement (RSI) and loss of control. The technology is cross-domain, while the state remains organized by domain. Ideally, the U.S. China bilateral AI dialogue itself will spur both sides to accelerate progress towards the type of institutional authority that the age of AI, RSI, AGI (artificial general intelligence), and ASI (artificial superintelligence) require, both domestically and in time, forming the foundation of a global organization that can tackle a minimally viable AI governance framework. For longer term discussions that go beyond the most pressing near-term CBRN safety concerns into RSI and other issues, it would also be helpful to review a recent paper “Beyond Rivalry” from CCA Sr. Fellow, Alvin W. Graylin, who provides a comprehensive framework for cooperative efforts to reduce risks around runaway AI, capex overbuild, model bias, economic instability and geopolitical equity.

The U.S.’s Frontier AI Model Governance System: Many Capabilities, No Settled Owner

The U.S. landscape can be read across five functions: technical evaluation, national-security review, industrial-policy consequences, diplomatic coordination, and political authorization. The problem is not that the United States lacks relevant institutions; it is that these functions sit in different places, and no agency yet combines all five. The graphical diagram of the key parties and their relationships below can help bring more clarity on the situation.

In the United States, relevant expertise across these areas of potential discussion is dispersed across multiple institutions, while authorities specific to the most frontier AI models, such as monitoring, evaluation, release power, are all basically non-existent. Because Treasury Secretary Scott Bessent has a long-standing interest in AI and has been given the authority by the President to lead the trade negotiations out of which the AI Dialogue has materialized, he will play an important role in managing the diplomatic and interagency dimensions of the U.S. China AI track, especially where AI security intersects with financial stability, cyber resilience, and broader economic-security policy. But the Treasury Department is not traditionally a major player in the AI domain within the U.S. interagency, so Bessent’s small staff will be stretched to coordinate with the many other players who will seek a seat at the table.

Cyber and security agencies: authority without the full frontier-model remit

The Office of the National Cyber Director (ONCD), the Cybersecurity and Infrastructure Security Agency (CISA), the National Security Agency (NSA), and the intelligence community illustrate the authority side of the gap. They bring cyber, classified, and threat-assessment capacity, but frontier-model governance extends beyond network defense into model behavior, deployment conditions, biosecurity, and commercial release decisions.

For example, the White House cyber apparatus and the Office of the National Cyber Director (ONCD), under National Cyber Director Sean Cairncross,^[4] has made a major power play to become the lead agency on frontier model governance, playing a major role in the drafting of the recent executive order on the topic. To be sure, ONCD brings cyber authorities and coordination capacity, but the Office is not a natural fit for leadership domestically, because the issue is bigger than just cybersecurity, and the natural focus of the organization is network defense, critical infrastructure, ransomware, and state-backed cyber operations. ONCD’s mandate centers on cyber coordination rather than frontier-model evaluation, and its institutional weight in this arena is limited relative to Treasury, which is close to the President, and to technically capable organizations such as NSA. Other players also involved in the process that led to the EO, specifically CISA and NSA, adding operational and technical expertise, including on defensive tools and classified cyber capabilities. The intelligence community assesses adversary capabilities and strategic threats. NSA has much greater bureaucratic heft than any of the other players and will insist on a major role in assessing frontier model capabilities and determining how and when they are released, given its dual offensive and defensive cyber mission. NSA clearly brings significant AI expertise to the game, but as a core intelligence organization is not necessarily a logical candidate to participate in the initial stages of the U.S. China AI Dialogue, until the scope and key issues become clearer.

Commerce, NIST, and CAISI: technical evaluation without regulatory control

The Department of Commerce, however, including NIST, remains central to standards, measurement, technology controls, and industrial policy. The NIST-subordinate Center for AI Standards and Innovation (CAISI), formerly the U.S. AI Safety Institute, is the single largest reservoir of frontier AI model evaluation and testing expertise within the U.S. government, and any serious effort to engage China on frontier model governance will need to include CAISI. OpenAI, Anthropic, Microsoft, Google/DeepMind, and xAI have all entered into arrangements allowing CAISI to evaluate advanced models. These agreements have strengthened the argument that CAISI should receive substantially greater funding because it is now expected to perform independent evaluations of systems costing billions of dollars to train. CAISI has been caught in the politics of frontier AI model governance. In April 2026 its newly appointed leader, former Anthropic researcher Collin Burns, was pushed out by the White House four days into the job over his prior stint with Anthropic, currently a controversial player in the sector after ongoing tussles with the Department of War and the former and still influential White House AI Czar David Sacks; the Commerce Department subsequently turned to Chris Fall, a former director of the Energy Department’s Office of Science, to lead the center.^[5] In June 2026, administration officials also reportedly directed CAISI to pause publication of its model evaluation reports while the new executive order is implemented, a move that underscores how contested the center’s public-facing role remains.^[6] The Commerce export control apparatus also has become involved with the recent order based on Export Administration Regulations (EAR) statutes, to bar use of Anthropic’s Fable 5 model by “foreign nationals” due to the potential for the model be subject to “jailbreaks” by foreign adversaries and non-state actors. This action may have highlighted the need for companies and national leaders around the world to ensure AI sovereignty, pushing them to evaluate Chinese open-source models that would be freed from such regulatory actions.

While NSA falls bureaucratically under the Defense Department, its director is also dual hatted as commander of U.S. Cyber Command, and the organization has a role in leveraging AI for military applications and reducing operational risk. OSTP and other White House and interagency actors are also involved in the emerging frontier-AI security framework. OSTP in particular, under Director Michael Kratsios, has been heavily involved on the AI promotion side, having a hand in drafting last summer’s U.S. AI Action Plan. Kratsios also maintains close relations with Undersecretary of State Jacob Helberg, whose Pax Silica initiative is designed to shore up the U.S. AI data center supply chain. While both Kratsios and Helberg could be candidates for participation in a U.S. China AI Dialogue given their bureaucratic remit, neither appear to have focused significant effort on the issue of governance of frontier AI models.

Mirroring the Chinese side to some degree, where some AI issues fall with the purview of the arms control elements of the Ministry of Foreign Affairs, the State Department Bureau of Arms Control, Deterrence, and Stability, under the Under Secretary for Arms Control and International Security is explicitly responsible for strategic stability, escalation risk, arms control, deterrence, and emerging-technology challenges, which makes it the natural home for the arms-control dimension of a U.S.–China AI dialogue.

But if the dialogue is more focused on laying the groundwork for bilateral agreement on scaffolding around the testing and release of frontier AI models, the arms control framework may prove to be unworkable, given the significant economic and commercial interests that are at play here. It is worth recalling that the 2024 Geneva round was not run by arms controllers on either side: the U.S. delegation was led by the NSC and the State Department’s special envoy for critical and emerging technology, while China fielded the MFA’s North American and Oceanian Affairs Department rather than its Department of Arms Control.^[7] Where each government chooses to house the new dialogue will itself be a signal of how it conceptualizes the problem.

The expertise-authority gap

The United States still has not decided where frontier-model evaluation belongs inside the government. Existing agencies can assess many related risks, but none clearly owns the question of when a model’s capabilities should trigger government review before deployment. CAISI has already been playing a significant role here but is not a regulatory body. Rather, it serves as a guiding organization which can make recommendations. This has produced an expertise-authority gap: some of the agencies with the strongest national-security authorities are not built around frontier-model evaluation, while the institutions with deeper technical expertise may lack political standing or decision-making power. This situation is rapidly changing in the wake of the June executive order, but there is still considerable churn within the administration of authorities and leadership in this rapidly changing arena.

CAISI is closest to the technical evaluation function, though its authority remains limited. It is not a licensing body and does not control model release, all its arrangements with the leading labs are voluntary. Its role is better understood as a technical bridge between frontier AI developers and parts of the national-security state. That makes CAISI indispensable for any serious discussion of model evaluation, including Chinese frontier models, but also politically awkward as a formal diplomatic counterpart: the same expertise that makes it useful for dialogue also ties it to sometimes politicized U.S. assessments of Chinese company frontier capabilities.

This points to a broader expertise-authority gap. Security agencies can evaluate adversaries, vulnerabilities, and strategic threats, but they are less directly built for model testing, including assessing emergent capabilities, interpreting benchmark results, or judging how deployment conditions affect risk. These are specialized capabilities that require regular interaction with the leading AI labs that currently only CAISI really has. The debate over giving CAISI a larger role in frontier-model review reflects this institutional gap.

One important pocket of capacity sits outside the cyber-centric framework altogether. The Department of Energy and its National Nuclear Security Administration (NNSA) are the only parts of the U.S. government that have actually run classified evaluations of frontier models as an ongoing practice: since April 2024, NNSA and the national laboratories have assessed Anthropic’s models for nuclear and radiological proliferation risk in a Top Secret environment, and the two sides have co-developed nuclear safeguards classifiers whose methodology is being shared with other developers through the Frontier Model Forum.^[8] If a bilateral agenda with Beijing eventually extends beyond cyber to the chemical, biological, radiological, nuclear (CBRN) risks both governments have flagged in official documents, China’s own AI Safety Governance Framework explicitly lists misuse in nuclear, biological, and chemical domains, then DOE and NNSA expertise will be hard to leave out of the room, and the CBRN lane may in fact offer the most tractable early agenda for technical exchange, since neither side wants proliferation-relevant capabilities diffusing to third parties. But there is an important distinction here between the C and the BRN. Recent revelations of Chinese cyber operations targeting U.S. critical infrastructure, distillation of U.S. frontier models, and broader and longer-term U.S. concern about China’s cyber theft of IP make this domain particularly sensitive to discuss in the context of AI-enabled offensive and defensive cyber operations.

The June 2026 executive order as a first test case

Complicating the issue is the fact that the June 2026 AI Executive Order establishes what is effectively the first U.S. government framework for classified evaluation of frontier AI models based on national security capabilities rather than general AI safety concerns. The centerpiece is a classified benchmarking process, led by national security agencies in coordination with CAISI, to assess advanced cyber capabilities and determine when a system qualifies as a “covered frontier model.” The benchmark itself would be classified, reflecting concerns that publicly disclosing capability thresholds could reveal sensitive information about offensive and defensive cyber operations. The framework is focused primarily on cyber-enabled risks, including vulnerability discovery, exploit development, and critical infrastructure impacts, rather than broader questions of misinformation, bias, or societal harms.

From a strategic perspective, the EO is attempting to move frontier AI governance closer to an arms-control and strategic stability model. Developers of covered frontier models are encouraged to provide the government with up to 30 days of pre-release access for classified evaluation, allowing agencies to assess whether emerging systems could alter the cyber offense-defense balance or create new national security risks. While participation remains voluntary, the EO creates an institutional mechanism for government visibility into the most capable AI systems before deployment. In effect, the United States is beginning to treat certain frontier AI models less like commercial software and more like strategically significant technologies requiring specialized government assessment, a development that could eventually inform future discussions with China on AI risk reduction, strategic stability, and confidence-building measures.

But this is very much work in progress, and many within the industry are uncomfortable with a classified evaluation process and will likely push back on this process. In addition, a further political difficulty is that even a voluntary review process can be seen by some actors as the beginning of a de facto licensing regime. That concern is especially strong when many policymakers view regulatory constraints through the lens of U.S.-China technological competition. At the same time, placing frontier AI too narrowly inside cyber-focused institutions could mischaracterize the problem. Cybersecurity is central, but frontier AI safety also raises questions around biosecurity, autonomous behavior, model control, misuse, and broader social risks. A process designed mainly for cyber defense may not capture the full range of frontier-model concerns. A broader model-release or licensing framework would be a later, even more involved step.

U.S. versus China: Industry and technical counterparts

Finally, the role of industry here in the process of setting up the dialogue remains complex. The expertise in frontier AI model development, testing, and deployment ultimately rests within small teams with all the leading labs. Any agreement between the two governments will have to establish a mechanism for inputs from both AI safety organizations such as CAISI and China’s CnAISDA, from other pools of expertise such as the Frontier Model Forum (FMF), and the labs themselves. CnAISDA, the China AI Safety and Development Association unveiled in February 2025 around the Paris AI Action Summit, deserves a precise read: it is not a CAISI-style government agency but a network of existing institutions, convened with backing from senior figures such as Turing Award winner Andrew Yao and Tsinghua’s Xue Lan, whose members include Tsinghua’s Institute for AI International Governance (AIIG), the Shanghai Qi Zhi Institute, the Beijing Academy of Artificial Intelligence, the Shanghai AI Laboratory (SAIL), CAICT, and the MIIT-affiliated think tank CCID. The network itself does not test or evaluate models, although several member institutions do.^[9] That asymmetry of institutional form, an agency on one side and a network on the other, could complicate any effort to designate formal technical counterparts, though it is likely that the U.S. China AI dialogue could force consolidation of frontier AI model testing capabilities into a smaller and empowered subset of CnAISDA member organization. Indeed, this would be a desired outcome. On the U.S. side, players such as Sacks could also be part of the mix, or better, whoever replaces him in the position of White House AI Czar^[10], should President Trump decide to appoint a suitable replacement, ideally someone who can straddle industry and government and help shape the development of the U.S. bureaucratic organization structure optimized for frontier AI model governance, and serve as bridge both between industry and the government, and for bilateral and multilateral discussions of the issue as a logical focal point that can reach out across the U.S. interagency for expertise and authority as required.

China’s Frontier AI Model Governance System: Emerging but Dynamic Ecosystem

As with the U.S. graphical diagram, the relevant Chinese parties and their relationships below should be studied closely to provide a visual understanding of the counterparts we will be working with.

China’s system can be mapped through the same five functions. CAC is central to online information, algorithm governance, and data security; MIIT to industrial policy and technical standards; NDRC to planning and infrastructure; MOST to research and innovation; MFA to external governance; MPS and MSS to cyber, public-security, and intelligence enforcement; and the PLA to military risk. The problem is that no single body specializes in frontier-model risk evaluation while also holding authority across these domains.

Major government actors: authority distributed across competing mandates

Ministries, administrations, and standards bodies in Beijing have issued a large volume of AI regulations and policy guidance, especially around algorithms, generative AI services, synthetic content, data compliance, online information control, and industrial development over the past three years. But like the U.S., Beijing has not chosen to establish an agency designed specifically to evaluate frontier AI risk. The CAC has authority over online information, public opinion, algorithm governance, content security, and data-related controls. MIIT is more closely linked to industrial policy and technical standards. MPS, through cyber and public-security channels, brings enforcement capacity over network security, cybercrime, and the security implications of deployment. NDRC’s expanding AI role points to planning, investment coordination, strategic resource allocation, and the integration of AI into broader economic policy. MOST remains relevant through science and technology policy, research funding, and innovation strategy, and a MOST Vice Minister has periodically represented China at international AI safety fora, including at Paris in 2025.

MFA has the external-facing function of engaging in global AI governance. That external engagement increasingly runs through Beijing’s preferred multilateral vehicles: the Global AI Governance Initiative of October 2023, the Global AI Governance Action Plan released at the July 2025 World AI Conference, and the proposed World AI Cooperation Organization (WAICO, to be headquartered in Shanghai, which Xi personally promoted at APEC in November 2025. Any bilateral dialogue will be shaped by Beijing’s effort to route AI governance through these China-led and UN channels. MSS and MPS are both centers of strong cyber and security expertise: MSS on intelligence and state-security risks, MPS on public-security enforcement, cybercrime, and protection of critical information infrastructure. The PLA has also been a reservoir of cybersecurity-related capabilities, and has clear stakes, given the cyber and military implications of frontier AI model capabilities, which have reportedly featured in recent military conflicts.

Chinese Technical Institutions

China also has an array of capable technical bodies. Institutions such as CAICT and emerging AI safety-focused organizations can provide practical expertise on testing, standards, and security assessment. The most concrete artifact of this technical layer is the AI Safety Governance Framework, issued in September 2024 and upgraded to version 2.0 on September 15, 2025, by TC260, China’s national cybersecurity standards committee, together with CNCERT/CC under CAC guidance. Version 2.0 explicitly addresses misuse in cyber and CBRN domains and loss-of-control risks and introduces a risk grading system, making it the closest thing China has to an official frontier-risk taxonomy and a natural reference document for any bilateral technical agenda.^[11] A handful of institutions are also building genuine evaluation muscle: SAIL, for example, under Director Zhou Bowen, published a Frontier AI Risk Management Framework with Concordia AI in July 2025; the Beijing Academy of Artificial Intelligence (BAAI), added to the Entity List by the Biden Administration in 2025, a status that would make its participation awkward; and the Beijing Institute of AI Safety and Governance, established in February 2025 under Chinese Academy of Sciences (CAS) scientist Zeng Yi.^[12] Zeng, who serves as a member of the United Nations High-Level Advisory Body on AI, is a widely respected AI safety expert who brings both technical expertise and multilateral engagement experience to the table. We will be watching for his inclusion in the Chinese team as a sign of how Beijing will approach the dialogue.

Research institutes linked to public-security, state-security, and cybersecurity systems may also be involved where frontier AI intersects with cyber operations, intelligence, and domestic-security concerns. The result is a policy field in which expertise, authority, and political access are distributed across different actors.

As in the United States, effective frontier-AI governance in China would require some combination of overall policy development, still lacking, coupled with technical competence, national-security authority, and access to top-level political decision-making. NDRC has gained influence over AI policy, but it is not primarily a technical agency and focuses primarily on domestic AI data center infrastructure and industrial policies related to the AI stack, not frontier model safety. MFA may become the visible face of China’s international AI governance efforts, while more powerful actors, including parts of the security apparatus, remain less visible. Technical institutions may be able to assess models, but they may not have the authority to define national-security priorities, and they do not appear, like CAISI, to have developed equivalent relationships with key expert bodies related to cybersecurity and biosecurity or with MPS-linked operational channels to enable testing of models under controlled or classified conditions. CnAISDA appears to have good relations with leading Chinese AI labs, but there does not appear to have been the deep integration of AI testers and evaluators within U.S. AI labs that CAISI has pioneered in the U.S. This reflects a familiar Chinese governance pattern: when a policy issue touches many priorities at once, many institutions can claim relevance, while ownership remains blurred.

Chinese AI governance must therefore reconcile competing mandates within a single political system. Frontier AI touches growth, security, ideology, diplomacy, military modernization, private-sector dynamism, and regime security at the same time. Internal coordination becomes politically and bureaucratically demanding.

This has direct implications for a Track 1 U.S.-China AI safety channel. A serious bilateral process will require counterpart capacity. Meaningful dialogue will be hard to sustain if the two systems lack functionally comparable institutional counterparts with enough authority and technical depth to scope and iterate a discussion of containing frontier-model risk. It also requires a shared understanding of what kind of risk is being discussed. If one side frames AI safety mainly as cyber defense while the other emphasizes access to systems for evaluation, the two sides can quickly talk past each other, making the collaboration framework moot.

This risk is already visible in debates over model access and evaluation. Chinese researchers have argued that meaningful safety assessment requires access to the systems being evaluated. In Washington, such arguments can easily be read as pressure for access to sensitive U.S. technology. Meanwhile, close cooperation between the U.S. government and leading AI firms can reinforce Beijing’s perception that American frontier labs are increasingly integrated into the national-security state. These perceptions make even technical exchanges politically charged.

Both sides will likely need new mechanisms that can evaluate emerging capabilities, define risk thresholds, and coordinate across multi agencies with competing mandates. Personnel continuity will also matter. Leader-level endorsement has elevated AI safety as a priority in the bilateral relationship, but it cannot substitute for institutional capacity to move the conversation forward.

History suggests that when transformative technologies emerge, governments often spend years building the bureaucracies needed to govern them. Nuclear technology eventually produced institutions such as the Atomic Energy Commission and later the Nuclear Regulatory Commission. Financial crises led to new regulatory structures after 2008. Frontier AI may now be at a similar stage.

Appendix: Who is Who in China’s AI governance system

China’s AI governance system is fragmented by design. Multiple agencies and political networks shape policy, often advancing overlapping agendas while checking one another’s authority. This pattern has become more pronounced since the 20th Party Congress, as Xi Jinping has balanced the portfolios of senior leaders rather than placing the full AI agenda under a single bureaucratic chain. The result is a coordination problem that will shape China’s response to frontier AI and limit what interlocutors can deliver in U.S.-China AI dialogues.

Formal negotiators may therefore not control the institutions that matter most. MOFCOM and MFA can participate in talks, but they do not have the domestic authority to align China’s military, security, industrial, and economic-planning systems. The key question is which officials and agencies can make frontier AI legible to Xi and influence his assessment of its opportunities and risks. Some of the most important players on AI may rarely interact directly with U.S. or multilateral counterparts but can still shape the advice and constraints facing China’s negotiators.

Politburo Standing Committee member Cai Qi could become a key player in the emerging U.S. China AI dialogue process. As perhaps the closest confidante of President Xi Jinping and the fifth-ranking member of the PBSC and director of the Central General Office, he is Xi’s chief of staff and one of the most important enforcers of ideological discipline, who also has responsibilities that touch on the risks and opportunities of frontier AI model deployments. His positions in the Central National Security Commission and the Central Cyberspace Affairs Commission, which sits over CAC, give him influence over the party-state’s information control and security apparatus.

Through the Central Cyberspace Affairs Commission, Cai has direct visibility into the Cyberspace Administration of China, the main regulator for online content, data security, and algorithm governance. The CAC has been led since 2018 by Zhuang Rongwen, who concurrently directs the commission’s general office and serves as a deputy head of the party’s Propaganda Department, a triple hat that ties AI regulation directly to the information-control apparatus. CAC’s filing and registration system functions as the primary clearance mechanism for generative AI services. It requires security assessments for developers of foundational large language models and works with technical bodies such as CNCERT/CC and emergency-response entities linked to MIIT and MPS. Because Cai sits above the cyberspace system, AI threat assessments generated through CAC and related security channels are likely to move quickly toward the top leadership.

Vice Premier Ding Xuexiang sits on the development and technology side of the system. As the sixth-ranking member of the Politburo Standing Committee and executive vice premier, he oversees macroeconomic planning, public finance, innovation policy, and the national science and technology agenda. Through the Central Science and Technology Commission, he helps coordinate major state-backed programs in AI, semiconductors, and data infrastructure. The commission is a strategic coordinator that identifies local policy experiments and scales selected models through the Ministry of Science and Technology and provincial party science-and-technology committees.

Vice Premiers He Lifeng and Zhang Guoqing occupy adjacent but important lanes. He, a Politburo member and vice premier, oversees finance, domestic commerce, monetary policy, and U.S.-China trade negotiations. His network also retains influence inside the NDRC, despite Ding’s formal portfolio. Zhang Guoqing, also a Politburo member and vice premier, manages the industrial base, informatization, and state-owned assets. Through MIIT and SASAC, he has influence over the physical backbone of China’s AI ecosystem, including computing infrastructure, industrial automation, hardware standards, and procurement rules affecting foreign AI chips in state systems.

The military operates through a parallel channel. The CMC Science and Technology Commission and its National Defense Science and Technology Innovation Rapid Response Groups monitor commercial breakthroughs and connect them to military users. Established in 2018, these groups function as local sensors for civil-military fusion, with visible activity in cities such as Chongqing, Xi’an, and Shenzhen. Despite recent PLA purges, including the March 2026 delisting of Liu Guozhi, who led the CMC Science and Technology Commission from 2016 to 2021, from the Chinese Academy of Sciences academician roster,^[13] open-source records suggest that these local nodes have continued to operate. This gives the PLA a direct mechanism for tracking frontier AI and passing military-relevant assessments upward.

The NDRC is emerging as another important coordination platform. Formally under Ding’s portfolio but also shaped by He Lifeng’s personnel network, the commission operates as a high-level strategy and evaluation body. Its AI and civil-military fusion work appears to rely on agile task forces that draw personnel from local development and reform commissions and other ministries. A reported Leading Group on Semiconductor Industry Development has its Office located at NDRC, and includes semiconductor component and an AI specific organization, managed by Xiangli Bin and Huang Ru, respectively. Huang is a microelectronics academician and NDRC vice chair; together these offices focus on semiconductor supply chains, hardware, computing infrastructure, and cross-agency coordination. The NDRC’s role in cases involving AI firms and overseas restructuring suggests that it is gaining leverage over the boundary between technology policy, capital control, and national security.

The National Data Administration, established under the NDRC in 2023, adds another piece to the system. Its mandate is to build a national data market and reduce the proprietary data silos controlled by large technology firms. That task is central to China’s AI ambitions, but implementation has been difficult. Firms have strong incentives to protect commercially valuable data, and early efforts to inventory corporate data assets have reportedly met resistance.

MPS is important for cybersecurity and public security. Through its 11th Bureau, the Cybersecurity Protection Bureau, it regulates enterprise network security under the graded protection system. As AI risks become more visible, MPS is increasingly focused on AI-related vulnerabilities and the security risks created by deployment. Its role overlaps with CAC’s authority over cyberspace governance and with the security services’ broader interest in cyber and intelligence risks.

China’s diplomatic interlocutors face clear constraints. MFA represents China in international AI talks and can frame broad principles on safety, sovereignty, and global governance. But it lacks the technical capacity, rank, and budget to direct the institutions that manage the AI ecosystem, including NDRC, CAC, MIIT, MPS, and the PLA. MFA also depends heavily on outside experts and think tanks, including the Tsinghua Institute for AI International Governance, led by founding dean Xue Lan, who also chairs the national New Generation AI Governance Expert Committee, with former vice minister Fu Ying as honorary dean.^[14] That gives it intellectual support, but not the in-house capacity or mandate to assess specific frontier-model risks.

MOFCOM has more direct policy leverage through export controls and trade tools. It can shape technology governance when AI intersects with external pressure, sanctions, or cross-border transactions. But its autonomy is limited. In sensitive cases, MOFCOM may initiate or manage parts of the process, but higher-ranking bodies such as the NDRC can still step in when strategic technology, capital, or national-security concerns are involved.

An expert ecosystem that can bridge the two systems is starting to form. CnAISDA knits together the institutions most engaged on frontier risk, and several of its convening figures carry considerable weight inside the system: Turing laureate Andrew Yao and Tsinghua AIR dean Zhang Yaqin have co-signed the International Dialogues on AI Safety red-lines statements alongside Western counterparts, Zeng Yi runs the Beijing Institute of AI Safety and Governance under the Chinese Academy of Sciences, and Zhou Bowen’s Shanghai AI Lab has published a frontier AI risk management framework with Concordia AI, the most active Track 2 connector between the two safety communities.^[15] None of these actors holds formal negotiating authority, but they are the likeliest source of the working-level technical expertise and trust a government channel will need, and any U.S. coalition strategy should map and engage this network deliberately rather than treating the official organs as the only interlocutors. So far, U.S. officials, including from CAISI, have not engaged seriously with China’s AI safety community, either at major international fora such as the Bletchley Park AI Summits or at other international AI fora where Chinese players have been present.

The challenge for U.S.-China AI dialogue is whether China can bring the above security, military, industrial, and planning systems into alignment, and which bureaucratic channels can translate frontier AI risks into language and priorities that reach top leadership.

[1] Center for Strategic and International Studies, analysis of the May 2026 Trump-Xi Beijing summit and its technology deliverables, May 2026, csis.org; PRC Ministry of Foreign Affairs summit readout, May 2026, fmprc.gov.cn.

[2] The White House, “Statement from NSC Spokesperson Adrienne Watson on the U.S.–PRC Talks on AI Risk and Safety,” May 13, 2024, bidenwhitehouse.archives.gov; PRC Ministry of Foreign Affairs, readout of the first meeting of the China–U.S. intergovernmental dialogue on artificial intelligence, May 15, 2024, fmprc.gov.cn.

[3] The White House, “Readout of President Joe Biden’s Meeting with President Xi Jinping of the People’s Republic of China,” Lima, Peru, November 16, 2024, bidenwhitehouse.archives.gov.

[4] Wall Street Journal reporting on implementation of the June 2, 2026, executive order, June 2026; The White House, “Promoting Advanced Artificial Intelligence Innovation and Security,” Executive Order, June 2, 2026, whitehouse.gov/presidential-actions.

[5] Washington Post, reporting on the removal of Collin Burns as CAISI lead and the selection of Chris Fall, April 24, 2026, washingtonpost.com.

[6] Wall Street Journal, reporting that CAISI was directed to pause public model-evaluation reports during executive order implementation, June 2026, wsj.com.

[7] See the White House NSC statement of May 13, 2024, and the PRC MFA readout of May 15, 2024, cited above.

[8] Anthropic, “Developing nuclear safeguards for AI through public-private partnership,” August 21, 2025, anthropic.com; Axios, reporting on the first frontier model evaluation in a Top-Secret environment beginning April 2024, November 14, 2024.

[9] Scott Singer, “How Some of China’s Top AI Thinkers Built Their Own AI Safety Institute,” Carnegie Endowment for International Peace, June 2025, carnegieendowment.org; Shanghai Qi Zhi Institute, readout of the CnAISDA unveiling on the sidelines of the Paris AI Action Summit, February 2025.

[10] Reuters and Bloomberg, reporting on David Sacks stepping down as White House AI and crypto czar at the 130-day special government employee limit and becoming PCAST co-chair, March 26, 2026.

[11] National Technical Committee 260 on Cybersecurity and CNCERT/CC, AI Safety Governance Framework 2.0, September 15, 2025, tc260.org.cn; Carnegie Endowment for International Peace, “How China Views AI Risks and What to Do About Them,” October 2025.

[12] Shanghai AI Laboratory and Concordia AI, Frontier AI Risk Management Framework, July 2025; Concordia AI, AI Safety in China updates on the establishment of the Beijing Institute of AI Safety and Governance under the Chinese Academy of Sciences, February 2025, concordia-ai.com.

[13] South China Morning Post, reporting on Liu Guozhi’s removal from the Chinese Academy of Sciences academician roster, March 19, 2026, scmp.com.

[14] Tsinghua University Institute for AI International Governance, institutional pages on its founding dean, honorary dean, and role supporting the New Generation AI Governance Expert Committee, aiig.tsinghua.edu.cn.

[15] International Dialogues on AI Safety, consensus statements (Ditchley Park 2023; Beijing 2024; Venice 2024), idais.ai; Concordia AI, State of AI Safety in China reports, 2023–2025, concordia-ai.com.

Old Wine, New Bottle: Why “Combat Effectiveness” Remains an Elusive Endeavor

You can learn a lot about PLA weaknesses simply from reading what they emphasize. In particular, three recent PLA Daily articles collectively illustrate a deliberate effort by political work departments within the PLA to reshape the values, training incentives, and leadership practices of the force. Read another way, their repeated emphasis on “combat effectiveness” and “political reliability” suggests training standards are not meeting expectations and ongoing purges in the PLA are having lasting reverberations throughout the military.

Although the articles focus on different services — including operational units, support organizations, and defense researchers — the common thread is that officers should prioritize long-term contributions to combat capability, organizational effectiveness, and national defense modernization over individual accolades.

Correcting Political Performance

The most prominent theme across the three articles is the need to establish a “correct political performance outlook” (政绩观). One article highlights an Air Force brigade explicitly warning against allowing short-term metrics to shape military activities, stating that personnel had developed a “utilitarian mindset that prioritizes evaluations and scores over practical utility” (重考核,轻实用的功利思想). Brigade leaders concluded that “simply staring at the current report card and not aiming at success on the battlefield of the future” (只盯着当前成绩单,不瞄准未来战场打赢) represented “a deviation in the view of political performance” (政绩观出现偏差). It added: “If not corrected in time, it will inevitably damage the combat effectiveness of the troops.”

Similarly, another article focused on the revolutionary PLA defense scientists who created China’s first ICBM and nuclear warhead, praising the older generation for “consciously making political achievements for the people and making political achievements with practical work” (自觉为民造福,以实干创政绩). It further encourages contemporary researchers to pursue achievements with the mindset that “I do not have to personally achieve success, but I must contribute to achieving success” (功成不必在我,功成必定有我).

“Combat Effectiveness” as the Central Standard

The three articles repeatedly emphasized “combat effectiveness” (战斗力) as the organizing principle for political work. The Air Force brigade article directly states that officers and soldiers must “resolutely correct ideological understandings and training styles that deviate from the combat effectiveness standard” (坚决纠正偏离战斗力标准的思想认识和训练作风). The brigade subsequently emphasized the principle of “how to fight, how to train” (怎么打仗就怎么训练), promoting realistic combat courses and developing complex opposition environments. One brigade officer summarized this approach: “Training is not for first place, but only for winning on the battlefield” (训练不是为了争第一,而是为了战场打赢). Likewise, an article about the party committee of an Information Support Force unit argues that resolving “long-standing administrative issues” allows officers and soldiers to “devote more energy to the main responsibility and main business of training and preparation for war” (把更多精力投入练兵备战主责主业).

Overcoming “Short-Termism” Through Party Committee Leadership

A third theme is an admonition not to pursue short-term behavior and rewards at the expense of long-term combat effectiveness goals, portraying party committees as essential actors in improving military effectiveness. The Information Support Force article credits the new party committee with taking historical problems as “an important annual task” (作为年度重要任务来抓) and maintaining continuity during leadership transition. The Air Force article describes party committee members conducting research, identifying shortcomings, and leading personnel to front-line training sites to solve difficult operational problems.

The same article offered this anecdote to underscore the leading role of party committees:

“For a period of time in the past, the brigade’s training courses did not match the actual combat requirements. At one meeting of the brigade, when the relevant departments of the organ and the major teams reported on the training situation, the members of the brigade party committee team raised a sharp question: the training content was old-fashioned and lacked breakthroughs, and new courses could not be added based on the requirements of the actual battlefield. In the face of the merciless questioning of the members of the party committee, the head of the training section of the brigade said frankly: “When formulating the training plan, there is indeed a preference for emphasizing evaluation and neglecting practical application; for the new courses that need to be tackled, due to the long effect cycle and the difficulty of quantifying the results, they are not included in the daily training plan.”

Finally, the article on missile scientists similarly connects political values and scientific innovation, praising earlier generations who “served the country wholeheartedly and worked tirelessly” (赤诚报国,忘我奉献). In each article, party organizations are presented as responsible for aligning personnel behavior with organizational goals and military requirements.

CCA Analysis: Taken together, these articles suggest that the PLA leadership remains concerned about organizational behaviors that can undermine military effectiveness, including what the PLA calls “careerism,” leadership turnover, bureaucratic incentives, and training practices disconnected from operational requirements. The recurring theme of “correcting deviations in political performance” and “focusing all minds and energy on combat effectiveness” indicates an effort to align individual incentives with the demands of military modernization.

The articles also imply that the PLA increasingly views professional competence and institutional responsibility as important components of political reliability. By linking political work to combat readiness and innovation, these narratives seek to shape the behavior of officers, commanders, and defense researchers as the PLA prepares for more complex operational missions and continued force modernization.

But all these traits — including seeking individual rewards and “box-ticking” training exercises — are deeply ingrained in the culture of the PLA. Clearly, the PLA is worried its lack of battlefield experience will impact its ability to simulate and achieve a high standard of “combat effectiveness” in training. What was noteworthy was the continued emphasis that party committees are taking the lead in holding the PLA to a higher standard. Given all the purges of senior officers in the Political Work department, one interpretation of this emphasis is that Xi Jinping and the CMC are worried about the image of the Political Work officers, especially political commissars.

Liaoning aircraft carrier task force completes one of the longest far-sea combat trainings on record

Chinese aircraft carrier Liaoning concluded 40 days of far-sea combat training and returned safely to its home port of Qingdao on Monday, June 22, the Chinese navy reported.

During the training, the task force operated in multiple sea and air spaces, including the South China Sea and the western Pacific, conducting carrier-based aircraft tactical flight drills and task force search and rescue exercises. The task force included the cruiser CNS Wuxi (104), destroyer CNS Kaifeng (124), frigate CNS Luohe (545), and fast combat support ship CNS Hulunhu (901). The Type 075 amphibious assault ship Anhui also took part in the exercises.

Footage was released appearing to show the first public video of a J‑15T carrier‑borne fighter jet being refueled in mid‑air from a PLAAF Yun‑20U tanker. Tanker refueling operations are a critical component of China’s blue water naval ambitions.

During the training, China’s Ministry of National Defense claimed that Japanese vessels and aircraft repeatedly conducted “close-range tracking and surveillance,” causing “disturbances and provocations.” The Global Times later released a video on X claiming to support China’s claim of harassment.

While China claimed the training was a “routine exercise in accordance with the PLA Navy’s annual plan,” Japan’s Maritime Self-Defense Force (JMSDF) reportedly shadowed the carrier group as it passed through the Okinawa and Miyako islands into the Philippine Sea, as it has done in the past in this strategically important waterway. Tokyo confirmed it monitored the fighter launches and movements, and later issued a statement that Japanese self-defense forces will “continue to conduct professional and steady vigilance in the surrounding sea and airspace of our country…under a calm and resolute stance.”

CCA Analysis: The current 40-day far-sea combat training deployment appears to be either the longest or tied for the longest continuously reported deployment in Liaoning’s operational history. Chinese state media explicitly emphasized the unusually long duration and the extensive combat-oriented nature of the training.

Several previous deployments approached this duration:

• December 2021: The Liaoning operated in the Yellow Sea, East China Sea, and western Pacific for roughly three weeks before returning to port.

• May 2022 western Pacific deployment: The carrier group conducted extensive operations near Japan’s southwestern islands and recorded more than 300 aircraft sorties over about 20 days at sea.

• September–October 2024 Joint Sword-related operations: The Liaoning participated in extended exercises around Taiwan and the western Pacific lasting approximately two to three weeks.

• June 2025 dual-carrier operations with Shandong: The Liaoning spent several weeks operating in the western Pacific alongside China’s second carrier, the Shandong, but publicly available reporting did not indicate a deployment exceeding one month.

Based on the publicly available record, this recent deployment is very likely the longest continuously reported operational deployment in Liaoning’s 14-year service history. It also appears to be the longest far-sea training deployment ever publicly acknowledged by the PLA Navy for any Chinese aircraft carrier. The PLA Navy rarely publishes precise deployment start and end dates, and some earlier deployments may have involved brief port calls or logistics stops that were not publicly disclosed. As a result, it is difficult to establish a definitive historical record of China’s aircraft carrier deployments.

Finally, regarding the accusation that Japan “interfered” with PLA operations, we have seen this movie before. The video evidence does not support the claim. In fact, my review of the footage suggests the Japanese military maintained a safe distance and operated professionally. This is more about Sino-Japanese politics than it is about operational safety.

CCTV Reveals New Details of the “Nantianmen Project” — China’s Futuristic Space Combat Force

On May 31, CCTV.com revealed the latest progress of the “Nantianmen Project” (南天门计划) – a future-oriented aviation-themed science fiction program under development since 2017. The project is a Chinese aerospace science fiction universe combining novels, videos, concept art, air-show exhibits, and full-scale mockups of futuristic military systems. It is an imaginative framework that integrates hypersonic flight, air-space vehicles, dual-mode aerospace propulsion, directed-energy weapons, metamaterial stealth, and AI-assisted autonomous swarms and drones into a future military ecosystem. The project is part of China’s long-term plan to construct a global space-based strategic defense system consisting of large strategic aerospace carrier platforms and aerospace fighters.

CCTV-7’s National Defense and Military Channel (国防军事频道) introduced new conceptual designs of the project, which included the 100,000-ton-class aerospace carrier “Luan Niao” (鸾鸟); the unmanned aerospace fighter “Xuan Nü” (玄女); a full-band stealth fighter “Bai Di” (白帝); and a universal vertical takeoff and landing platform “Zi Huo” (紫火).

Screenshot of the “Luan Niao” aerospace carrier. Source: CCTV News

The project aligns with Shanghai’s “15th Five-Year Plan” strategic deployments regarding the digitalization of cultural tourism, commercial aerospace, artificial intelligence, and intelligent robots. The project recently opened an exhibit in the Lingang New Area aerospace industry cluster in Shanghai, creating China’s first immersive cultural tourism science fiction park based on aviation science fiction IP. Lingang New Area is at the forefront in positioning Shanghai as a globally influential science and technology innovation center.

The Nantianmen Project was created by AVIC Global Culture & Technology (Beijing) Co., Ltd., a subsidiary of the SOE conglomerate Aviation Industry Corporation of China (AVIC). AVIC envisioned using real-world technology to construct a science fiction world for the future aerospace era.

The three previously unveiled concept fighter jets differ from modern fighter jets already in existence in terms of aerodynamic layout, drive engine, flight mode, and payload.

The centerpiece of the universe is the “Luan Niao,” an enormous “aerospace carrier” capable of operating in near-space and launching multiple subordinate aircraft. It is often depicted as a strategic command platform. It is 100,000 tons, 242 meters long, has a wingspan of 684 meters, and is capable of carrying 88 “Xuan Nü” fighter jets.

The “Xuan Nü” fighter jet is conceptually designed as an unmanned air superiority fighter. It features meson circulation generators on both sides of the fuselage, and is capable of using particle acceleration cannons, hypersonic missiles, and other weapons. Combined with its forward-swept wing design, it has a visually striking effect and can operate outside the Earth’s atmosphere.

The “Bai Di” full-band stealth fighter jet employs full-band stealth and can switch between manned and unmanned flight modes. It uses a variable-wing structure to adjust its aerodynamic layout in real time according to altitude, speed, and flight attitude.

Finally, the “Zi Huo” is a universal vertical takeoff and landing platform featuring a sharply defined fuselage, a force field duct, and a tiltable power nacelle. A prototype was unveiled at the 7th China Tianjin International Helicopter Exposition in October 2025.

“Zi Huo” multi-purpose vertical takeoff and landing platform. Source: Global Times

CCA Analysis: None of these capabilities have been demonstrated in reality. For the PLA, the Nantianmen Project is therefore best interpreted as a window into China’s aspirational aerospace future rather than a catalog of actual PLA capabilities. Nonetheless, the project provides a venue for Chinese engineers, designers, and science-fiction enthusiasts to conceive of future aerospace warfare without committing to specific procurement programs, and many concepts could influence real research decades later. Chinese state media have explicitly framed the project as a way to popularize aerospace science and inspire young people to pursue careers in aviation and defense technology.

Many of the systems, like the “Luan Niao,” are almost Star Wars-esque in their rendering with a Chinese twist. The appearance of full-scale mockups at major air shows inevitably blurs the line between fiction and future capability. Some Chinese analysts believe this helps generate discussion about China’s long-term aerospace ambitions, even if the systems themselves are fictional.

For analysts of Chinese military modernization, the Nantianmen Project is useful because it offers insight into how components of China’s aerospace-industrial community envision warfare in the 2040–2060 timeframe. Many of the themes align with real-world PLA and Chinese aerospace research priorities, such as the integration of air and space operations and distributed swarm technology.

China Launches Naval and Coast Guard Drills Near Taiwan

In the past week, China has deployed both the China Coast Guard (CCG) and the PLA Navy to signal opposition to expanding Japan–Philippines security cooperation, particularly Tokyo and Manila’s decision to begin maritime boundary delimitation talks in waters that overlap with maritime areas near Taiwan that are claimed by Beijing.

Between June 1 and 7, China’s Ministry of Transport (MoT) announced a “Special Maritime Traffic Law Enforcement Action” and for the first time dispatched MoT vessels alongside CCG vessels to patrol east of Taiwan. The “law enforcement patrols” explicitly linked the operation to the Japan–Philippines maritime boundary initiative, which Beijing described as an “infringement” on China’s maritime rights.

Taiwan subsequently reported increased Chinese maritime activity in waters southeast of Taiwan, including near Orchid Island (Lanyu) and the strategically important Pratas (Dongsha) Islands. Chinese coast guard vessels repeatedly entered waters claimed by Taiwan around Pratas, leading to multiple confrontations with Taiwan’s Coast Guard Administration. One confrontation included a tense exchange between the Taiwanese and Chinese coast guard in which a Taiwan coast guard officer relayed the following message over the radio: “China Coast Guard, these are not your waters. You don’t belong here. Turn around and leave now!”

Screenshot of a China Coast Guard social media post tracking the movements of CCG vessel 2502 east of Taiwan. Source: China Coast Guard Via Weibo.

Around the same time, the PLA Navy maintained a higher operational tempo in the western Pacific. Japan reported that the aircraft carrier Liaoning and its escorts conducted extensive flight operations east of the Philippines, including roughly 170 aircraft and helicopter sorties, demonstrating China’s ability to operate beyond the First Island Chain.

CCA Analysis: The delimitation talks are problematic for Beijing because they implicitly treat waters east of Taiwan as an area that can be legally negotiated between Japan and the Philippines without Chinese participation. China has therefore used Coast Guard patrols east of Taiwan to reinforce its claim that Taiwan-related maritime zones fall under Chinese jurisdiction. The choice of the CCG rather than an overt PLA Navy operation is consistent with China’s broader “gray-zone” approach of asserting sovereignty and jurisdiction while remaining below the threshold of a major military confrontation.

The Chinese response appears calibrated to send three messages. First, to show Japan and the Philippines that Beijing rejects any maritime arrangements that could strengthen legal or operational cooperation in waters it claims. Second, to show Taiwan that China is increasingly using the Coast Guard to normalize its presence around Taiwan-controlled islands and challenge Taipei’s jurisdiction. This is the first time MoT and CCG vessels have conducted coordinated patrols east of Taiwan, and it represents a new form of MoT-CCG integration near Taiwan. Finally, it sends a message to the broader region that carrier operations east of the Philippines and coast guard patrols east of Taiwan demonstrate China’s ability to apply pressure simultaneously in the Taiwan Strait, the South China Sea, and the western Pacific.

From a PLA-watching perspective, the noteworthy trend is the growing integration of CCG “law enforcement’ operations with PLA military signaling, allowing Beijing to apply pressure below the threshold of major military exercises while steadily expanding its operational presence around Taiwan.

We’re glad to have you here for the inaugural issue of the China Energy Brief, a monthly newsletter from the China Climate Hub. Each issue offers forward-looking, evidence-based insights into the forces shaping China’s green tech and energy action—and what it means for the U.S. and beyond.

This month, we take you onto the floor of the world’s largest-ever Beijing Auto Show, where the future of the electric vehicle industry was on full display. Beyond the gleaming new models, we unpack two trends with far-reaching implications: the accelerating global push of China’s cleantech exports, and how those tailwinds are reshaping Beijing’s leverage in its dealings with the rest of the world.

We hope you find this first issue valuable. If this isn’t the kind of content you signed up for, no hard feelings—you can unsubscribe from China Energy Brief in your Substack subscription preferences, using the button below, and tailor the content you receive from us:

Adjust Subscription Settings

Otherwise, we look forward to bringing you our analysis each month.

Enjoy the read.

China Energy Brief: Issue 1

Written by Kate Logan, Director of the China Climate Hub

The Beijing Auto Show was held from April 24 to May 3, setting an official record as the world’s largest auto show by both scale and vehicle count. What was once a standard car exhibition has morphed into a showcase for the forefront of electric vehicle innovation. I visited this year’s show, which coincided with the lead-up to Trump’s recent summit with Xi in Beijing and took place as the global energy crisis from the Iran conflict entered its third month. Beyond the shiny new car models, two important trends reflect how China and its leading firms are approaching their global strategies.

First, China’s cleantech exports will grow even stronger, which will pivot cleantech companies’ hypercompetitive mindsets into the global domain.

The official tagline of the Beijing Auto Show was translated as “future of intelligence.” But a secondary, unofficial theme permeating countless booths was “made in China, for the world”—underscoring the shift from serving the domestic market to serving overseas consumers. This trend is driven by both domestic and international factors.

Within the Chinese market, demand for EVs and other cleantech is waning, especially as incentives are removed and prices and purchases adjust accordingly. Weakening demand started “pushing” Chinese cleantech exports toward global markets well before the current energy crisis. One analyst shared that they had to revise their earlier projections that EV exports would slow by mid-2025 after observing sustained strong flows, especially in markets with few trade barriers, such as Australia and the Philippines. Export growth has been further amplified in recent months by the “pull effect” from international consumers looking to take advantage of low-cost clean energy amid spiking fuel prices. In April, domestic auto sales fell for the seventh straight month and were down 21.6% year-on-year, while EV and plug-in hybrid exports surged to more than double the same month a year prior.

Amid the Auto Show’s 17 massive exhibition halls spanning two separate conference centers, nearly every major automaker showcased a new, competitively priced luxury electric or hybrid-electric SUV—a trend that surprised me given the international obsession with China’s uber-affordable electric sedans, such as the BYD Seal. The factors elevating SUVs to the forefront reflect shifting market dynamics. Some of these vehicles will be sold domestically in an attempt to recover declining revenues. But with many of these models appearing nearly identical, many brands emphasized their desire to court global consumers, especially in high-end markets, including the Middle East. In Europe, hybrid vehicles are not subject to the same tariffs as EVs, paving the way for these battery-electric hybrids to make greater inroads.

While the show was about cars, batteries were the real stars. Attendees clustered around the pavilions of CATL and BYD’s battery arm, both of which featured cold chambers to showcase their batteries’ consistent performance even in Arctic environments ranging from –30 to –50 degrees Celsius (approximately –22 to –59 degrees Fahrenheit). While BYD emphasized the speed of its flash chargers—“ready in 5, full in 9”—CATL’s centerpiece was battery swapping, with a working model demonstrating the speed and convenience of swapping. CATL has touted its use of large language models (LLMs) to revolutionize battery chemistry research, thus freeing up revenues to spend on other areas of its global strategy.

The technologically sleek, all-electric vehicles of Chinese firms stood in contrast to those of several legacy foreign automakers, such as BMW, Audi, and Mercedes-Benz, which positioned some of their most historic vehicles as the centerpieces of their pavilions—as if to remind consumers that they had invented the internal combustion engine that powered the original automobile revolution.

Second, the tailwinds behind China’s exports have empowered Beijing to wield its leverage in responding to various demands from other countries.

In the weeks before the Trump-Xi Summit, letters emerged from House Democrats and Republicans, respectively—followed by bipartisan House and Senate bills—urging President Trump to prevent Chinese EVs from entering the U.S. market. Legislators feared he might strike a deal with Xi, akin to Canadian Prime Minister Mark Carney’s offer, in which Canada would import a limited number of Chinese EVs in exchange for Chinese investments into Canada.

But few in Beijing were closely following each escalation of this American pushback. Instead, Chinese players are eyeing more compelling opportunities elsewhere. This dynamic is perhaps also underscored by the Chinese companies invited by Xi to the state banquet with Trump during the summit: the overall absence of China’s cleantech darlings, save for perhaps tech brand Xiaomi, reflects Beijing’s hesitation to push deals if Washington remains ambivalent about accepting them.

The lingering effects of China’s October 2025 export controls on rare earth elements and lithium-battery materials still loom large, with companies lacking confidence that bilateral relations are stable enough to guarantee that Beijing will renew its temporary suspension. Moreover, Chinese companies that had been pursuing deals involving international tech transfer in other markets are facing down bureaucratic uncertainty from their own government. As one example, Chinese battery maker Hithium’s proposed partnership with Indian renewable conglomerate Reliance has yet to move forward despite sustained efforts. The important role of China’s Ministry of Commerce (MOFCOM) as a closed-door arbiter introduces significant procedural uncertainty into potential deals; new State Council regulations on overseas investment will further compound this. Beyond bureaucratic factors, competition among leading cleantech firms with similar products is so intense that preserving leverage vis-à-vis rivals may further dampen companies’ willingness to share their technology.

Commercial interests are most attracted to those markets that have issued some sort of political signal that Chinese firms are welcome to invest. As one example, while cynicism abounds toward the EU’s proposed Industrial Accelerator Act—especially since Japanese and Korean firms would be exempt from the various strings attached to future Chinese investments—enthusiasm for the Spanish market remains strong in the wake of Spanish Prime Minister Pedro Sánchez’s trip to Beijing. Following the visit, SAIC Motor announced its decision to locate a new EV factory in Spain rather than Hungary, underscoring how political signals can tilt Chinese investment decisions even at the margins. Reports also indicate that wind manufacturer Ming Yang, fresh off a UK rejection on security grounds, is now eyeing the Spanish market.

These two trends will no doubt accelerate global electrification—but they will also pose tough questions for international markets, especially those struggling to preserve their local industries. As I wandered around the show, I could not help but wonder how long global consumers could resist the high quality and marked affordability of the products I was seeing. Will other markets continue doubling down on trade barriers, or will they reconsider them? And will Beijing be willing to engage with increasingly restrictive investment environments, or will it start abandoning them in lieu of those that welcome Chinese firms?

Leave a comment

1. At Shangri-La, China Saves Its Fire for Japan

What Happened: From May 29 to 31, world leaders and defense officials from about 45 countries gathered in Singapore for the Shangri-La Dialogue, Asia’s premier security summit. The forum has historically given Beijing a platform to articulate its regional security positions, but this year’s dialogue was shaped less by direct U.S.-China confrontation than by broader concerns over strategic mistrust, rising defense spending, military modernization, and lessons from Ukraine.

Why it Matters: For the second year in a row, Chinese Defense Minister Dong Jun skipped the forum, signaling Beijing’s reduced interest in a venue where it is often scrutinized over its regional behavior. While China’s delegation criticized “certain countries” for “hegemonism” and “bloc confrontation,” the overall U.S.-China exchange was tame. U.S. Defense Secretary Pete Hegseth said bilateral ties were “better than they’ve been in many years” and avoided direct mention of Taiwan in his prepared remarks. The sharper exchange came instead between Beijing and Tokyo: China accused Japan of reviving militarism and questioned whether it could “win the trust of the international community,” while Japan criticized China’s military expansion and warned against changing the status quo by force. The contrast suggests that, even as Beijing manages optics with Washington after the recent Trump-Xi summit, it remains willing to pressure U.S. allies — especially Japan — over regional security issues.

By Kristina Lozinskaya, Schwarzman Fellow and Junior Fellow on China-Russia Relations at the Center for China Analysis

Learn More: Read “PLA Watch,” a monthly newsletter centered on delivering insights into China’s military affairs on the Center for China Analysis Substack.

View this email in your browser.

The latest on what’s happening in China from Asia Society Policy Institute’s Center for China Analysis.

THIS WEEK:

China targets Japan at Shangri-La, Beijing tightens overseas investment rules, Xi touts youth exchanges, policymakers advance urban renewal, and Xi spotlights future industries.

Read back issues in the China 5 Archive.

Police officers stand on patrol during the 23rd IISS Shangri-La Dialogue at the Shangri-La Hotel on May 29, 2026 in Singapore. (Photo by Ezra Acayan/Getty Images)

1. At Shangri-La, China Saves Its Fire for Japan

What Happened: From May 29 to 31, world leaders and defense officials from about 45 countries gathered in Singapore for the Shangri-La Dialogue, Asia’s premier security summit. The forum has historically given Beijing a platform to articulate its regional security positions, but this year’s dialogue was shaped less by direct U.S.-China confrontation than by broader concerns over strategic mistrust, rising defense spending, military modernization, and lessons from Ukraine.

Why it Matters: For the second year in a row, Chinese Defense Minister Dong Jun skipped the forum, signaling Beijing’s reduced interest in a venue where it is often scrutinized over its regional behavior. While China’s delegation criticized “certain countries” for “hegemonism” and “bloc confrontation,” the overall U.S.-China exchange was tame. U.S. Defense Secretary Pete Hegseth said bilateral ties were “better than they’ve been in many years” and avoided direct mention of Taiwan in his prepared remarks. The sharper exchange came instead between Beijing and Tokyo: China accused Japan of reviving militarism and questioned whether it could “win the trust of the international community,” while Japan criticized China’s military expansion and warned against changing the status quo by force. The contrast suggests that, even as Beijing manages optics with Washington after the recent Trump-Xi summit, it remains willing to pressure U.S. allies — especially Japan — over regional security issues.

By Kristina Lozinskaya, Schwarzman Fellow and Junior Fellow on China-Russia Relations at the Center for China Analysis

Learn More: Read “PLA Watch,” a monthly newsletter centered on delivering insights into China’s military affairs on the Center for China Analysis Substack.

2. China Expands Oversight of Overseas Investment

What Happened: Beijing issued sweeping new rules on June 1 that expand regulators’ authority over overseas investment involving Chinese investors, technology, data, and national security. The State Council regulation requires approval for overseas transfers of export-controlled goods, technologies, services, and related data, and bars indirect transfers through cross-border personnel deployment, consulting, and training arrangements. Building on the National Development and Reform Commission’s 2017 outbound investment rules, the new framework extends oversight to a broader range of overseas interests acquired directly or indirectly by firms, organizations, and individuals.

Why It Matters: The changes signal a shift from traditional outbound investment controls toward a more integrated national security regime. While earlier frameworks primarily targeted corporate outbound investments, the new rules also extend explicit regulatory reach to individual investors and non-traditional channels, especially with respect to sensitive technologies and data. This arrives as Beijing has already launched a crackdown on offshore trading platforms accused of helping retail investors bypass capital controls, while regulators have also moved against brokerages like Tiger and Futu over illegal cross-border securities operations.

By Shengyu Wang, Research Assistant, Center for China Analysis

Learn More: Watch “Are Export Controls Working? Field Notes from China’s AI Frontlines,” with Scott Kennedy, Trustee Chair in Chinese Business and Economics at the Center for Strategic & International Studies, and Paul Triolo, CCA Honorary Senior Fellow on Technology, moderated by CCA Fellow Lizzi C. Lee.

3. China Sets National Framework for Urban Renewal

What Happened: The 15th Five-Year Plan for Urban Renewal was released on May 28, turning a patchwork of local policy experiments into a national framework for long-term urban redevelopment. The plan reflects Beijing’s effort to move beyond the era of rapid outward city expansion and toward upgrading existing urban stock: improving older housing, pursuing adaptive reuse of old districts and factories, strengthening public services, and balancing development with cultural-heritage protection. It also explicitly warns against “large-scale demolition and construction” and “fake reconstruction,” marking a shift away from earlier redevelopment models that frequently relied on tearing down low-value areas for higher-density commercial projects. The key financing change is that eligible urban-renewal projects can now use local-government special-purpose bonds as project capital.

Why It Matters: Beijing is trying to redirect developers, construction demand, and local-government investment away from the land-financing model that contributed to China’s property crisis. The plan’s financing architecture also shows how difficult that shift will be. It combines special bonds, policy-bank lending, REITs, public-private partnerships, and utility-pricing reforms to make projects viable while avoiding new hidden local debt. If it works, urban renewal could become a more disciplined model for upgrading aging cities and infrastructure without fully reviving land-finance dependence. If returns on new projects disappoint, however, Beijing risks creating another channel for local debt.

By Jeremiah May, Research Assistant, Center for China Analysis

Learn More: Read “China’s Property Rebalancing: The Long Road to a New Development Model,” by CCA Fellow Lizzi C. Lee and David Zhang, Analyst at Trivium.

4. Xi Jinping on Developing China’s Future Industries

What Happened: In the June 1 issue of Qiushi, Xi Jinping published “Looking Ahead to the Future and Developing Industries,” an article based on his January 30 speech during the 24th collective study session of the 20th CPC Central Committee Political Bureau. The article calls for developing future industries through innovation and strategic planning, with a focus on sectors such as quantum technology and biomanufacturing. It also highlights the importance of enterprises, supportive policies, governance, and international cooperation in advancing China’s long-term development goals.

Why It Matters: Xi’s speech and Qiushi article signal China’s strategic priorities for achieving technological self-reliance and economic competitiveness in emerging industries. By emphasizing frontier sectors alongside stronger enterprise participation and policy support, Beijing is positioning future industries as key drivers of innovation, industrial upgrading, and long-term national development.

By Taylah Bland, Fellow on Climate and the Environment, Center for China Analysis (@Taylahbland)

Learn More: Follow China’s DeepSeek Moment, CCA’s webinar series examining key moments in China’s technological rise and their implications for global competition.

5. Xi Highlights Early Success of U.S.-China Youth Exchange Initiative

What Happened: Xi Jinping responded to a letter from students participating in a U.S.-China youth exchange program, highlighting that since he launched the initiative in November 2023 to “invite 50,000 American youth to China for exchanges and study over five years,” they have exceeded this goal and achieved the target two and a half years early. Xi noted that these exchanges have deepened mutual understanding, fostered lasting friendships, and opened a new chapter in people-to-people ties between China and the United States.

Why It Matters: The “50,000 in Five Years” initiative was announced during Xi’s 2023 visit to California for his meeting with then-President Biden, marking his first trip to the United States since the pandemic. Supported largely through programs such as the Young Envoys Scholarship (YES) initiative, the effort was initially met with skepticism regarding both its feasibility and strategic intent. Its completion well ahead of schedule nevertheless demonstrates that when top leaders on both sides actively support engagement—and neither side erects significant political obstacles—certain areas of U.S.–China cooperation can transcend partisan divisions and outperform expectations. Going forward, it will be important to watch whether Beijing and Washington continue to prioritize youth exchanges as a means of sustaining long-term people-to-people ties and cultivating a new generation of bridge-builders across the Pacific.

By Lobsang Tsering, Senior Research Associate on Chinese Politics, Center for China Analysis

Learn More: Read “Global Perceptions of China: Insights from the Next Generation,” featuring four winning essays from emerging young scholars around the world on how China is viewed from their corners of the globe.

Subscribe now

1. Xi Reportedly Planning North Korea Trip

What Happened: A possible visit by Xi Jinping to North Korea is being actively discussed by regional governments and media outlets. According to a May 20 Reuters report citing South Korea’s Yonhap News Agency, Xi could travel to Pyongyang in late May or early June. Reuters also reported that Chinese security and protocol teams had recently visited Pyongyang, often a sign of preparations for a high-level state visit. The Chinese government has not officially confirmed the trip, however.

Why it Matters: The proposed trip would be highly significant for China–DPRK relations. It would be Xi’s first visit to North Korea since 2019, following Xi’s recent summits in Beijing with Donald Trump and Vladimir Putin. The visit could also signal a shift in China’s approach toward “denuclearization” of the Korean Peninsula. China continues to publicly endorse “denuclearization” in principle, but recent statements suggest a growing emphasis toward “stability,” “dialogue,” and opposition to “deterrence and pressure.” Finally, the trip could help Beijing shore up its influence over Pyongyang. Beijing has arguably lost leverage since Vladimir Putin and Kim Jong Un signed a strategic partnership treaty in June 2024 that includes a mutual defense provision widely interpreted as a de facto mutual defense pact.

By Lyle Morris, Senior Fellow on Foreign Policy and National Security, Center for China Analysis (@LyleJMorris)

Learn More: Read the roundtable report from “The China–Russia–North Korea Nexus,” a closed-door discussion convened by the Sungkyun Institute of China Studies (SICS) and the Asia Society Policy Institute’s Center for China Analysis.

2. Beijing Tightens Expectations for Advisors Through Revised Responsibility Guidelines

What Happened: Wang Huning, a Politburo Standing Committee member and chairman of the Chinese People’s Political Consultative Conference (CPPCC), recently chaired a meeting calling on CPPCC members to strengthen research and consultation around key tasks set by the CCP Central Committee. Wang urged members of the Party’s political advisory body to offer policy recommendations that are forward-looking, targeted, and practical. The meeting also approved revised guidelines on “Strengthening the Responsibility and Commitment of CPPCC Members.”

Why It Matters: The specific revisions have not yet been disclosed. But when the guidelines were first introduced in 2020, their core message was that CPPCC members—who serve as part of the Party-state’s advisory and policy consultation apparatus—should “be neither magpies nor crows, but only woodpeckers.” In practice, this means they are expected neither to exaggerate positive news nor amplify pessimistic narratives, but instead to identify concrete problems and propose workable solutions. This principle has become an important guideline for Chinese scholars and policy experts, many of whom are CPPCC members, when engaging with overseas think tanks and foreign interlocutors.

By Lobsang Tsering, Senior Research Associate on Chinese Politics, Center for China Analysis

Learn More: Read “Guided Autonomy: America Watching in Xi’s China” by CCA Senior Fellow Guoguang Wu.

3. China Expands Social Insurance Access for Migrant Workers

What Happened: On May 22, China’s State Council announced new measures to encourage migrant workers to enroll in social insurance programs where they work, regardless of their hukou—China’s household registration system. The move affects China’s 357 million migrant workers, including gig economy workers on ride-hailing and food delivery platforms. The reforms aim to simplify cross-regional insurance transfers, with officials framing the changes as a way to help narrow the rural-urban gap, accelerate urbanization, and boost consumption.

Why It Matters: The move is part of China’s broader effort to dismantle the hukou system. To boost domestic consumption, Beijing recognizes the need to equalize access to public services, including education and healthcare, that remain tied to residents’ place of household registration. The reform also addresses mounting pressure on China’s social insurance system, which faces rapid population aging, sluggish contribution growth, and uneven participation among migrant and informal workers under the existing pay-as-you-go framework. Extending access to these workers could expand the social insurance system’s contribution base and improve its long-term fiscal sustainability.

By Shengyu Wang, Research Assistant, Center for China Analysis

Learn More: Read “Who Will Till the Land? Conflicting Narratives About Rural China’s Demographic Future,” by Vivianne Zhang Wei, founder of Chinese Farm Chronicles, and CCA Fellow Lizzi C. Lee.

4. Li Qiang Highlights Energy and Food Security in Zhejiang Visit

What Happened: Chinese Premier Li Qiang visited several strategic storage and logistics facilities in Zhejiang province, including the Zhoushan National Petroleum Reserve Base, a bulk agricultural product storage and transportation base, and the Ningbo Daxie commercial petroleum reserve project. During the tour, Li emphasized the need to maintain adequate reserves of “bulk commodities and important materials” amid growing uncertainty in the external environment. He also called for expanding reserve capacity, accelerating storage infrastructure construction, improving technological and digital management systems, and strengthening coordination between strategic and commercial reserves.

Why It Matters: While China’s reserve expansion is not new, the political signaling behind Li’s visit is significant. Senior Chinese leaders do not choose inspection stops randomly, and the focus on petroleum and grain reserves underscores Beijing’s growing emphasis on economic security and resilience in a more unstable geopolitical environment. The call to better coordinate strategic and commercial reserves also suggests China is building a more comprehensive resilience system to withstand supply disruptions and political shocks. More broadly, the visit reflects how Beijing increasingly views energy and food security as core components of national security.

By Li Shuo, Director of China Climate Hub, and Senior Fellow, Center for China Analysis

Learn More: Read “Why Food Security is a Top Priority for China” by former Research Associate Genevieve Donnellon-May.

5. Deadly Shanxi Mine Explosion Exposes Safety Failures

What Happened: On May 22, a gas explosion occurred at the Liushenyu mine in Qinyuan County, Changzhi, Shanxi. After initial uncertainty over the death toll, authorities now say 82 people are dead, two are missing, and more than 120 have been hospitalized. Xi Jinping and Li Qiang have framed the disaster, now under State Council investigation, as a warning for the country amid broader efforts to improve workplace safety.

Why It Matters: Reporting by Chinese state media has pointed to poor business practices and systematic rule evasion by the mining company, Shanxi Tongzhou Group. These include the company’s failure to maintain an accurate count of workers underground, as well as its outsourcing of labor to contractors under arrangements that weakened accountability and prioritized production over safety. Some reports also point to unregulated mining areas omitted from official maps, including the use of so-called “yin-yang drawings,” in which one set of drawings was submitted for regulatory filings while another guided actual mining operations. Following the incident, public commentary on Chinese social media has emphasized how catastrophic industrial accidents reflect long-term failures in enforcement and organizational cultures that normalize risk. There is also concern that local authorities initially minimized the death toll, reporting only eight deaths.

By Michael Cerny, Affiliated Researcher, Center for China Analysis

Learn More: Read “China’s New Climate Targets Decoded: Cautious Progress or Missed Opportunity?” by CCA Senior Program Officer Taylah Bland.

Subscribe now

Recent Southern Theater Command Exercises Highlight New Joint Training Thinking

A PLA Daily article, published on May 23, highlights ongoing efforts within the PLA to pursue joint, “real combat” training criteria over individual accolades. Of course, “jointness” is not a new push within the PLA, but the specifics of this report reveal new dimensions, and struggles, in honing a truly joint force.

The PLA Daily reporter follows several recent joint training exercises in the Southern Theater Command (STC), in particular involving the PLA Navy (PLAN) and Air Force (PLAAF). During one exercise, a pilot reportedly locks onto enemy aircraft, ready to fire, but the “real-time data-sharing and integrated battlefield awareness” (数据链实时共享战场态势一体感知) detects that friendly forces have not yet completed their attack deployment. The pilot assesses that launching an attack would “prematurely expose their own position, disrupt the entire combat plan, and even lead to the failure of the system-wide operation.” (提前暴露己方位置，打乱整个作战部署，甚至导致体系作战失利). The report says the pilot immediately abandons his attack and disengages to coordinate with friendly forces, concluding that “the system’s power created a formidable synergy, allowing them to quickly gain the upper hand on the complex and ever-changing battlefield” (体系威力形成强大合力，让他们很快在复杂多变的战场上占据主动).

The takeaway, according to the after-action report, is as follows:

“In the past, we always focused on individual air combat performance and the pass rate of individual subjects; but when we actually went to the joint training ground, we realized that no matter how excellent an individual is, it is difficult to form a cohesive force without the system; no matter how outstanding the performance is, it is difficult to achieve effective training without coordination.” (过去, 我们总盯着单机空战成绩, 单课目优秀率; 可真到联合训练场上才明白: 个体再优秀, 脱离体系也难成合力; 成绩再突出, 没有协同也难有练兵实效).

The unit’s leader explained that every officer and soldier should have this sense of the “big picture” in mind when training within joint simulated environments.

The article went on to describe the ongoing challenge for troops and officers in changing their mindset from seeking individual accolades and scores to adopting “joint force” standards. The author repeatedly mentioned instances of pilots and officers sacrificing personal rewards for the betterment of the joint force. Or, as the article says, “pooling their collective wisdom around the principles of ‘unified command, unified discourse, and unified action’” (统一指挥, 统一话语, 统一行动集智攻关), and “breaking away from traditional evaluation models to adopt the system contribution rate as the main standard for measuring training effectiveness” (打破传统考评模式，将体系贡献率作为衡量训练质效的主要标准).

CCA Analysis: This is a useful reminder that the PLA has a long way to go to achieve a true joint force. The examples cited in the article about progressing step by step from joint training with the “same type of aircraft,” to “mixed training with different types of aircraft,” and then to “integrated ship-aircraft training” are illuminating. For the U.S. joint force, many of these scenarios would seem rudimentary (granted, the U.S. military has by no means “perfected” joint operations; it is always a work in progress). It is also illuminating to read what a massive mindset shift is required for the PLA to train under true joint requirements and environments. This is a historically army-based military which, as the article concedes, has relied on individual exercise scores that likely inflate their sense of accomplishment. For all the handwringing these days in Washington that the PLA is 10-feet tall and outpacing the United States in several net assessment metrics, it still struggles with adopting a joint force mindset and accurately assessing training in joint settings. To me, RAND’s 2015 report on “China’s Incomplete Military Transformation” still applies to the PLA in 2026.

“Comradeship Should Not Become a Loophole in Discipline”: PLA Discipline Organs Target Officers’ “Social Circles”

The PLA is in the midst of one of the largest purges of senior officers in recent history. This is no surprise to anyone who has followed recent developments involving Zhang Youxia and Liu Zhenli, for example.

But a recent PLA Daily article identifies a new campaign undertaken by Party Committees and Discipline Inspection departments within the PLA — cracking down on “personal social circles” among PLA officers, especially while off duty.

The article, titled “An Army Unit Strictly Regulates the ‘Social Circles’ of Its Leading Cadres,” follows a “leading cadre” who recently took leave and received several phone calls from former comrades and family members asking for favors. The article reads as both a cautionary tale and a training guide, warning PLA cadres of new regulations imposing restrictions on meeting with and doing favors for family and friends while on vacation.

The scenarios themselves are revealing: one involves a call from a former comrade-in-arms who invites the officer out to drink; another from a family member who requests the officer’s help getting his nephew a civilian position in a local PLA unit; and a third from a manager of a local construction company.

The first scenario paints the following picture: a retired army veteran reaches out, asking to get together. He says, “It’s been a long time since we’ve seen each other. You’re the only one still in the army. Let’s get together and catch up!” The officer responds, “Feelings are feelings, and discipline is discipline. Our camaraderie doesn’t need to be maintained by eating and drinking. Let’s play a game of badminton tomorrow morning, work up a sweat, and chat. Wouldn’t that be great?” The friend agrees and, after badminton, asks the officer, “Does your company need to purchase training equipment recently? I have resources and connections.”

Upon hearing this, the officer waves his hand and says, “Lao Liu, if you’re going to talk about this, then that’s enough for today. You know the regulations in the army. We’re comrades-in-arms, but there’s no room for personal relationships when it comes to rules. Don’t make me make a mistake.”

The second scenario involved a phone call with a relative, during which a military officer was asked to use his connections to help a nephew secure a civilian position in another military unit. The cousin urged him to “put in a good word” and ask recruiters to “‘take care’ of him.” The officer firmly refused, explaining that the recruitment process is now “entirely supervised by the discipline inspection department” and that every stage is “open and transparent.” When the cousin expressed disbelief that he lacked enough “influence” after years in the military, the officer responded that “this isn’t about saving face; it’s about the bottom line of discipline.” He emphasized that cadres are prohibited from using their authority to benefit relatives and concluded that if the nephew truly wanted the position, he should “prove himself based on his abilities.”

The last scenario starts when the officer receives a call from a local construction company manager inviting him to “relax” at a newly opened fishing park, insisting they “absolutely would not talk about work.” The officer recognizes the caller as a manager from a company that had recently lost a bid for a military repair project. Rejecting the overture, the officer bluntly explains that the company failed in its bid not because it lacked “connections in the military,” but because “your quote and proposal were indeed inferior to those of the winning bidder.” Recalling a corruption case in which a PLA official was gradually compromised through seemingly harmless leisure invitations, the officer stressed that PLA engineering projects are now “completely open and transparent,” with the Discipline Inspection Commission supervising “the entire process.” He added that “if the conditions are met, there’s no need to treat anyone to meals or entertainment,” before ending the call.

CCA Analysis: This article is super interesting for several reasons. First, the three scenarios reveal sources of corruption that have long plagued the PLA for decades: family members and business colleagues using “guanxi” networks within the PLA to enrich themselves. Zhang Youxia was reportedly accused of failing to control “close associates, family members and relatives” tied to graft networks, and Li Shangfu and Wei Fenghe were also accused of abusing their authority over procurement within the Central Military Commission (CMC) Equipment Development Department in exchange for financial gain and political favors. There has also been speculation that China’s nuclear silo lids not operating properly as a result of embezzlement.

More noteworthy is the fact that the CMC Discipline Inspection Commission (中央军委纪律检查委员会) has apparently launched new “anti-corruption education classes” throughout the PLA, including having Party Committee Secretaries and Deputy Secretaries to conduct “one-on-one reminder talks” with members of the unit leadership teams on “hunting” scenarios involving family members and business associates. To quote the article, “purify your ‘three circles’ (自觉净化 “三圈”) (i.e. circles of influence, social circles, and personal relationships), and that the use of power must be “based on evidence and interactions must leave a trace.” (用权需有据, 交往应有痕).

The article is almost comical in its admonition for officers not to speak with or attend social gatherings with family friends or close associates during leave. But it’s no laughing matter: this is one of the most common avenues for corruption to take place in the PLA. Good luck enforcing it.

Liaoning Departs for Another Exercise Near Japan

On May 19, the PLA dispatched the Liaoning — China’s first aircraft carrier — for a new round of naval exercises in the western Pacific that included live-fire drills, tactical flight operations, and support missions, according to the Ministry of National Defense website. The PLA Navy described the drills as “routine annual exercises” intended to improve combat readiness and operational capabilities.

A day later, Japan’s joint staff tweeted that its JASDF Southwestern Air Defense Force’s fighters scrambled to interdict suspected PLA intrusions into Japan’s airspace over the East China Sea.

The PLA Navy deployment follows a series of increasingly ambitious Chinese naval operations beyond the first island chain, including previous exercises east of the Miyako Strait and near Guam. Last year, Chinese and Japanese aircraft were involved in a tense encounter near Okinawa during a similar carrier deployment, with Japan accusing Chinese J-15 fighters of locking fire-control radar onto Japanese aircraft.

CCA Analysis: China is still seething over Japanese Prime Minister Sanae Takaichi’s public remarks that a Taiwan Strait crisis could justify Japanese military intervention, comments Beijing condemned as crossing a “red line.” China is sending the PLA to remind Tokyo of Beijing’s indignation. China has noticeably expanded unilateral military activity around Japanese territory, with Japan’s latest defense white paper warning that growing Chinese military operations in nearby waters and airspace pose a threat to Japan’s security. Sino-Japan relations have reached the lowest level in years, and these latest carrier drills are China’s attempt at normalizing sustained naval and air operations in the western Pacific near Japan.

1. Beijing Defines “Constructive Strategic Stability” After Trump-Xi Summit

What Happened: The most important new framing to emerge from the Trump-Xi summit was the idea of “a constructive China-U.S. relationship of strategic stability.” Beijing has elevated this as the summit’s defining outcome, with similar language echoed by the White House. Chinese officials and state media are presenting it as guidance for the next three years and beyond, even elevating it as a major Xi-era innovation in great-power relations. The concept moves beyond Cold War-style strategic stability rooted in rivalry and mutually assured destruction. Instead, Beijing’s official interpretation stresses expanding shared interests, avoiding zero-sum competition, preventing sharp swings in ties, and steering clear of conflict, confrontation, or war.

Why It Matters: Beijing’s motivation behind “constructive strategic stability” reflects both confidence and anxiety. On one hand, China wants this framework to reduce external volatility amid China’s domestic economic weaknesses, elevate Xi as the central architect of leader-level diplomacy, and position China as a historic power shaping a new paradigm with the United States. On the other hand, the push also reflects mounting pressure stemming from China’s economic slowdown, geopolitical headwinds, uncertainty surrounding the Iran crisis and resulting global turmoil, and fears that unmanaged competition with Washington could harden into confrontation, or worse. The appeal of this concept is that it allows China to seek guardrails while positioning itself as an equal to the United States, while preserving room for cooperation in critical areas such as public health and AI safety, even as strategic competitive fundamentals intensify.

By Lizzi C. Lee, Fellow on Chinese Economy, Center for China Analysis (@wstv_lizzi)

Learn More: Explore “The Avoidable War: An Alternative Future for U.S.-China Relations,” an ASPI initiative on alternative futures for U.S.-China relations, and read “U.S.-China 21: The Future of U.S.-China Relations Under Xi Jinping,” Kevin Rudd’s framework for managing strategic divergence and cooperation within a stable long-term relationship.

2. April Data Shows China’s Uneven Recovery

What Happened: April data revealed the fragility of China’s domestic recovery. Retail sales slowed to just 0.2 percent year-on-year, fixed-asset investment fell back into contraction, private investment weakened, and the property slump showed few signs of easing. Exports remained a key cushion, but their strength was increasingly concentrated in tech and high-end manufacturing—sectors that boost output and headline growth but do not generate enough broad-based employment or income support to revive household confidence.

Why It Matters: The Trump-Xi summit may have temporarily eased tensions, as reflected in MOFCOM’s readout on bilateral trade, but China’s deeper challenge lies at home. Households are borrowing less, spending cautiously, and deleveraging amid a weak labor market and prolonged property downturn. Private firms, especially smaller firms in traditional sectors, are being squeezed from both sides: weak domestic demand limits pricing power, while higher energy prices linked to the conflict in Iran are raising input and supply chain costs. China’s tech engine is buoying sentiment, but it is not pulling the wider economy or the labor market along with it.

By Shengyu Wang, Research Assistant, and Lizzi C. Lee, Fellow on Chinese Economy, Center for China Analysis (@wstv_lizzi)

Learn More: Read “China’s Property Rebalancing: The Long Road to a New Development Model,” by CCA Fellow Lizzi C. Lee and David Zhang, Analyst at Trivium.

3. Xi Hosts Putin Five Days after Trump in a Bid for Reassurance

What Happened: Five days after President Xi held a high-stakes summit with President Trump, Beijing welcomed President Putin with comparable grandeur. Commemorating the 30th anniversary of the “China-Russia strategic partnership of coordination,” Xi praised their strategic cooperation as reaching a “historic high,” while Putin hailed the two as a stabilizing force in the world. The leaders signed around 40 agreements focused on the economy, trade, education, science, and technology.

Why It Matters: Putin’s 25th visit to China, coming on the heels of Xi’s historic meeting with Trump, is meant to reassure Moscow of its continued importance to Beijing while projecting a united front amid lingering tensions with the United States after the Trump-Xi summit. In an indirect jab at Washington, the two leaders warned of the unprecedented damage caused by unilateralism and hegemony. They criticized the war in Iran as a violation of international law, called the United States “irresponsible” for allowing the New START nuclear treaty to expire, and condemned Trump’s proposed Golden Dome missile defense system as a threat to global stability. At the same time, the visit revealed limits in the bilateral relationship. The two sides failed to reach a breakthrough on the Power of Siberia 2 pipeline, which would have doubled Russia’s natural gas exports to China, underscoring Beijing’s reluctance to deepen long-term dependence on Russian gas despite mounting concerns over energy security. The Chinese foreign ministry also refuted reports that Xi had told Trump that Putin might regret his 2022 invasion of Ukraine, raising questions about the extent of China’s support for Russia in the war.

By Kristina Lozinskaya, Schwarzman Fellow and Junior Fellow on China-Russia Relations, Center for China Analysis

Learn More: Watch “China’s Role in and Perception of a Post-Ukraine Conflict Scenario,” with CCA Non-Resident Senior Fellow Ambassador Pierre Andrieu and CCA Senior Fellow Lyle Morris.

4. Record EV Exports Sustain China’s Cleantech Momentum

What Happened: Analysis of monthly data from China’s General Administration of Customs indicates that China’s cleantech exports in April 2026 remained close to record highs, despite changes to tax rebate policies that were expected to reduce solar and battery exports. Although solar and battery exports declined, they remained considerably above pre-Strait of Hormuz crisis levels. At the same time, a surge in electric vehicle (EV) exports drove monthly export volumes to a record USD 9.4 billion, more than offsetting the decline in solar and battery exports.

Why It Matters: Against the backdrop of ongoing geopolitical tensions and instability in global energy markets, the April data provides the clearest evidence yet that disruption to the Strait of Hormuz is accelerating demand for Chinese clean technology products. This is especially visible in the EV sector, where electrification offers an alternative to dependence on traditional internal combustion engines. Continued strength in overall cleantech exports also suggests that demand has remained resilient despite policy adjustments aimed at reducing solar and battery export volumes.

By Taylah Bland, Fellow on Climate and the Environment, Center for China Analysis (@Taylahbland)

Learn More: Read “The Evolving Politics of Climate Change in China“ by CCA Senior Fellow Guoguang Wu and CCA Fellow Neil Thomas.

5. Xi Calls for Accelerating China’s Independent Discourse and Knowledge System

What Happened: On May 17, Xi Jinping issued instructions on promoting the development of China’s philosophy and social sciences at a symposium. He called for deeper systematic and academic study and interpretation of the Chinese Communist Party’s innovative theories, while accelerating the construction of an independent Chinese knowledge system in philosophy and social sciences. Separately, more than 200 representatives from China’s social science community gathered in Beijing for a two-day symposium to study and implement Xi Jinping Thought on Culture.

Why It Matters: In order to promote its own discourse system and value standards, the CCP has historically sought to create new governing concepts, propositions, and theoretical frameworks, such as the concept of “One Country, Two Systems.” Since Xi came to power, Beijing has accelerated these efforts. In politics, Xi introduced the concept of “whole-process people’s democracy” while in economics, he promoted the idea of “new quality productive forces.” More recently, during his meeting with Trump, Xi proposed a “constructive relationship of strategic stability” between China and the United States, extending the same impulse to shape how China’s role is described in the international arena. Through these formulations, Beijing seeks not only to shape international narratives, but also define the terms on which China is understood.

By Lobsang Tsering, Senior Research Associate on Chinese Politics, Center for China Analysis

Learn More: Read “Guided Autonomy: America Watching in Xi’s China” by CCA Senior Fellow Guoguang Wu.

Subscribe now

Over the past 14 months, the Trump administration has presided over a significant loosening of export controls on AI chips to China – including the relicensing of H20 and H200 chips, and the establishment of a revenue-sharing agreement between NVIDIA and the U.S. government.

At each step, proponents of relaxation have insisted these chips are “deprecated,” a generation or more behind the cutting edge, and maintained that denying them accomplishes nothing while China builds its own alternatives. It is better, they claim, to keep Chinese developers hooked on NVIDIA’s software ecosystem — such as CUDA and related libraries and optimization tools — than to force Chinese AI developers into using Huawei’s Ascend chips and prompt China to build its own technology stack.

A growing chorus of industry-aligned analysts and consultants has amplified this logic into a broader indictment of the United States’ broader project to control technology exports to China. The “small yard, high fence” doctrine, they argue, has expanded beyond its initial remit — provoking retaliatory Chinese restrictions on rare earths, damaging U.S. chipmakers’ revenue and R&D capacity, and accelerating the very Chinese self-sufficiency it was designed to prevent. The Asia Society’s Paul Triolo, for example, has advocated lifting controls altogether on memory chips, redefining “advanced” upward to exclude everything but the absolute frontier, and effectively returning to the pre-2022 status quo.

This diagnosis gets the problem exactly backward.

China’s “Acceleration” Is Not What It Appears

The most commonly held argument for loosening export controls is that they have backfired — either by accelerating China’s domestic chip industry by forcing investments in self-sufficiency that Beijing might otherwise have deferred; or by forcing Chinese companies to work begrudgingly with Huawei when they otherwise would not have.

This narrative overstates both the pace and the quality of Chinese substitution in the AI industry. China has pursued semiconductor self-sufficiency since 2014, spending hundreds of billions of dollars and squandering billions more in waste, fraud, and abuse. It is important to understand that U.S. export controls did not create the Chinese government’s interest in indigenization — they forced Chinese firms to adopt inferior alternatives during a critical window of the industry’s development, with material impact to China’s hardware and software industry. SMIC’s 5nm-class chip fabrication process is estimated to be 40 to 50 percent more expensive than TSMC’s equivalent, with yields reportedly around one-third of TSMC’s on the same node.

Chinese developers are already beginning to feel the effects of this widening compute moat. Successfully forcing Chinese AI companies to turn to inferior hardware or foreign rental during a period of explosive AI growth is not a failure of export controls; it is direct evidence of the strategic friction the controls were meant to impose. Directly selling H200 — or better — chips to China directly undercuts the intended frictions that previous controls imposed.

Moreover, China’s AI chip fabs face component shortages far beyond lithography equipment. They compete against NVIDIA and AMD for access to the very same TSMC fabrication and packaging lines that produce the chips Washington is trying to deny. As things stand, Chinese chipmakers are not capable of supplying enough high-bandwidth memory or advanced packaging to build large numbers of AI accelerators on pace with the United States. This means that selling completed NVIDIA modules to Chinese AI labs hands them chips they could never have fabricated themselves, assembled from materials they cannot yet produce at scale, on production lines where their own designs are queued behind American orders.

Providing Chinese AI labs with more compute only strengthens their capacity to train and run inference for globally competitive AI services, while also freeing up Huawei’s scarce compute for potential export along a “Digital Silk Road.” Selling scarce chips directly to Chinese developers strengthens their ability to compete against U.S. companies in global markets.

There Has Never Been a More Important Time to Restrict China’s Access to Compute

Critics of U.S. technology restrictions are correct to point out that, despite being made to rely on a smaller quantity of less-performant hardware, Chinese AI labs have succeeded in building globally competitive models with a fraction of the compute available to their American counterparts. DeepSeek’s R1 rivaled frontier American systems at a fraction of the training cost; and Chinese open-source models — led by Qwen and DeepSeek — now account for nearly thirty percent of global open-source AI usage.

But there is a world of difference between building a model and serving it to the planet. The latter is the crux of the diffusion race U.S. and Chinese AI labs today find themselves locked into. Even though H200 exports to China will primarily serve domestic AI workloads, the absolute increase in compute will allow Chinese AI companies to more readily serve their products to global publics — undercutting the Trump administration’s aspiration to export the American AI stack.

Chinese labs continue to face critical compute shortages. Zhipu AI, which released its open-weight GLM 4.7 model in December, was forced to cap new sign-ups for its coding product to 20 percent of prior daily registrations after demand overwhelmed its servers. DeepSeek logs a steady drumbeat of service degradation events — the telltale signs of an inference fleet straining under load.

Proponents of relaxing U.S. export controls have consistently described the Trump administration’s decision to resume licenses for H20 and H200 as harmless because they trail the cutting edge by a generation. This misunderstands what the chips are for. Chinese AI companies do not necessarily need Blackwell-class GPUs to serve inference at global scale; they need a prodigious volume of decent compute to keep their APIs responsive and their prices low as they expand into markets from the Gulf to Southeast Asia to Latin America. Chinese AI executives themselves identify computational power as the one bottleneck constraining their global competitiveness. The correct response to that admission is to tighten the pressure, not ease it.

A Higher Fence Prevents the Yard from Expanding Further

If there is a legitimate criticism of the current U.S. technology control framework, it is not that the yard is too large but that the fence is too short. The $2.5 billion Supermicro indictment exposed an uncomfortable truth long known to U.S. export control enforcement officers: The current U.S. compliance system, which runs on paper declarations, end-use certificates, and the good faith of intermediaries stretching across dozens of jurisdictions, can be defeated through a method as simple as relabeling shipments in a rented warehouse.

The Chip Security Act, which the House Foreign Affairs Committee advanced on a bipartisan basis this week, could address this problem. Location verification for exported chips does not expand the yard, but raises the fence for protected technologies. Such an approach is fundamentally pro-diffusion: if the United States can confirm that chips shipped to Abu Dhabi or Singapore have not, in fact, been rerouted to Shenzhen, the case for loosening license requirements for other global markets is made materially stronger.

America’s technology siege is working as intended. The task now is to maintain it – and reinforce it where the fence has gaps — long enough for American compute, software, and APIs to embed in global markets, before Chinese chipmakers become capable of contesting the global AI market on their own terms.

Leave a comment

China Sentences Two Former Ministers of Defense to Death with Reprieve

On May 7, Xinhua News reported that a PRC military court sentenced the last two disgraced PRC Ministers of Defense, Wei Fenghe (魏凤和) and Li Shangfu (李尚福), to death with a two-year reprieve for corruption. Below is a rough translation of the article:

“On May 7, 2026, the military court lawfully pronounced judgment on the bribery case of Wei Fenghe, former member of the Central Military Commission, former State Councilor and Minister of National Defense. Wei Fenghe was found guilty of bribery and sentenced to death with a two-year reprieve, deprivation of political rights for life, and confiscation of all personal property. After the two-year reprieve period expires, his sentence will be commuted to life imprisonment in accordance with the law, and he will be imprisoned for life without the possibility of commutation or parole. On May 7, 2026, the military court lawfully pronounced judgment on the bribery case of Li Shangfu, former member of the Central Military Commission, former State Councilor and Minister of National Defense. The court found Li Shangfu guilty of bribery and offering of bribes, and sentenced him to death with a two-year reprieve, deprivation of political rights for life, and confiscation of all personal property. After the two-year reprieve period expires, his sentence will be commuted to life imprisonment in accordance with the law, and he will be imprisoned for life without the possibility of commutation or parole.”

CCA Analysis: The swift sentencing of Wei and Li marks a dramatic downfall for the two former CMC members. China rarely sentences CMC members — let alone former ministers of defense — to death. They are the first known cases of Chinese ministers of defense receiving death sentences with reprieve.

In 2023, Air Force General Liu Yazhou (刘亚洲) was given a suspended death sentence for corruption. Liu was formerly the Political Commissar of the PLA National Defence University and Political Commissar of the Chengdu Military Region Air Force. The last corruption-related scandal of a former CMC member was for former CMC Chief of the General Staff Fang Fenghui (房峰辉), who was purged in 2017 and received a life sentence in 2019. Before Fang, the last PLA general sentenced to death was former lieutenant general Gu Junshan (谷俊山), who received a suspended death sentence in August 2015 for corruption-related crimes. Gu served as the deputy director of the PLA General Logistics Department (GLD) from December 2009 to February 2012.

One also recalls the recent cases of former CMC Vice Chairs Guo Boxiong (郭伯雄), who received life imprisonment in 2016 for corruption, and Xu Caihou (徐才厚), who died of cancer in 2015 before trial proceedings concluded on charges of corruption.

The last major upheaval of this proportion occurred in the wake of the Lin Biao (林彪) purge in the 1970s. General Huang Yongsheng (黄永胜) reportedly received a death sentence with reprieve during the “Lin Biao clique” purge by Mao, later commuted; and Lieutenant General Wu Faxian (吴法宪), also reportedly sentenced to death with reprieve after the Lin Biao affair, later commuted.

For Xi to hand out suspended death sentences, Wei and Li likely engaged in other serious crimes beyond corruption, like leaking secrets or harming national security.

China Discloses Fifth-Generation Fighter Export Variant

During a May 1 Labor Day special program aired on China Central Television (CCTV), footage showed a J-35 stealth fighter with the serial number “0001” and Aviation Industry Corporation of China (AVIC) markings emerging from a hangar. The aircraft appeared without PLA Air Force insignia and was presented in an unpainted light grey finish, suggesting it may be an export variant of the J-35, called “J-35AE.”

Chinese military commentator Fu Qianshao said that the configuration — specifically a single-wheel nose landing gear rather than the reinforced dual-wheel system typically associated with carrier operations — indicates a land-based variant rather than a naval carrier-based model. The serial number “0001,” together with AVIC branding on the fuselage, further suggests a distinct production line separate from PLA operational inventory. In fact, months ago, the AVIC had already displayed a scale model of the J-35A at the 2026 Singapore Airshow, signaling export positioning in advance of the CCTV appearance.

So far, Pakistan has been widely assessed as a likely customer. In earlier days, Pakistani officials publicly stated that China had offered to sell 40 J-35 aircraft alongside KJ-500 airborne early warning platforms and HQ-19 air defense systems. On May 7, 2026, a senior Pakistan Air Force official further confirmed that an initial collaborative mechanism for the J-35AE had been signed, though contract details were not disclosed.

CCA Analysis: The emergence of the J-35AE in public signals that China is moving forward with a fifth-generation fighter export market. If confirmed at scale, this would represent a notable shift in the global stealth fighter landscape, where the United States has cornered the operational export market for fifth-generation fighters (The Russian SU-57 is excluded as there is no verified information or public disclosure of buyers so far).

It should be noted that the picture of the J-35AE was embedded within a broader state enterprise showcase, rather than a formal export announcement. Thus, some ambiguity remains over whether the disclosure was primarily informational, promotional, or deliberate signaling. Even so, the parallel confirmation from Pakistani officials of an initial collaborative agreement, combined with design features consistent with export configuration, indicates an institutionalized pathway toward foreign acquisition, with Pakistan as the most likely recipient.

More broadly, the development reflects how China’s defense export model is increasingly evolving beyond platform delivery toward integrated support arrangements. The reported presence of AVIC technical personnel in Pakistan during the May 2025 India-Pakistan crisis suggests a model in which Chinese systems are accompanied by embedded operational and technical support under real conflict conditions. Arms sales to Pakistan, including the recent delivery of the Hangor-class attack submarine, provide China with opportunities to expand influence and test its operational and support capabilities on the battlefield.

At the same time, it remains uncertain whether such an export program can scale in a manner comparable to the U.S. F-35 ecosystem. The U.S. model is anchored in alliance structures, long-term interoperability frameworks, and security guarantees that China currently does not yet replicate. While China may be entering the stealth fighter export space, its ability to turn this into a durable global network remains unproven.

The PLA Showcases Counter-Drone Air Defense Capabilities

China continues to publicly showcase its counter–unmanned aerial system (C-UAS) capabilities. According to CCTV in early May, the PLA Navy recently completed a “finalization test” (定型试验) of a new terminal air defense and anti-missile weapon system in the Bohai Sea. The test involved multiple drone targets simulating ultra-low-altitude sea-skimming profiles and low-observable penetration tactics under complex electromagnetic conditions. The test was designed to stress the system’s interception performance in realistic threat environments. The term “finalization test” indicates that a late-stage verification process in which a system that has largely completed development is moving toward production and deployment.

Chinese state media also publicized directed-energy counter-UAS systems. A recent CCTV program presented simulated engagements involving the “Guangjian-11E” (光箭-11E) and “Guangjian-21A” (光箭-21A) laser systems. The Guangjian-11E was shown in a “soft-kill” (软杀伤) role, disrupting drone sensors and communication links through pulsed laser effects, while the Guangjian-21A was demonstrated in a hard-kill (硬杀伤) role, physically damaging airframes and internal electronic components. The demonstrations included engagements against low-altitude targets and fiber-linked drones.

CCA Analysis: The development reflects a broader shift in which counter-UAS capability is increasingly treated as a core requirement of modern force design rather than auxiliary air defense. Operational experience from Ukraine and the Middle East has demonstrated that small, low-cost drones can generate disproportionate battlefield effects through reconnaissance, targeting support, saturation attacks, and one-way strike missions against high-value platforms. This has prompted the PLA to adapt layered air defense systems utilizing counter-UAS, particularly in maritime and littoral environments.

The PLA’s emphasis on ultra-low-altitude sea-skimming (超低空掠海), concealed penetration (隐蔽突防), and complex electromagnetic environments (复杂电磁环境) reflects concern over drone employment concepts designed to exploit radar coverage gaps. These characteristics are particularly relevant in a Taiwan contingency environment, where U.S. and Taiwan defense discussions increasingly emphasize a drone-enabled “hellscape” concept in the Taiwan Strait, intended to increase operational cost and delay PLA maneuver and landing operations.

From a capability development perspective, the PLA systems being demonstrated suggest an emerging layered counter-UAS structure combining soft-kill and hard-kill directed-energy options, typified by the Guangjian platforms. China’s development of high-power microwave (HPM) systems, including the “Hurricane-3000” (飓风-3000) we previously covered, offer another example of the PLA’s interest in counter-swarm applications. Compared to lasers, HPM systems are generally better suited to area effects, with the potential to disrupt multiple drone systems simultaneously through electromagnetic interference.

However, compared to laser-based systems, publicly observable evidence of operational HPM field employment remains limited. This may indicate that while the capability is being developed and tested, it has not yet reached the same level of operational maturity in the PLA. U.S. development programs such as the Tactical High-power Operational Responder (THOR) and its follow-on Mjölnir similarly suggest that HPM systems face persistent constraints in power generation, thermal management, beam control, and system integration under operational conditions.

Leave a comment