What the Trump-Xi Summit Revealed, and Left Unsaid, About U.S.-China Tech Competition

Last week, President Trump met with Chinese President Xi Jinping in Beijing, marking the first visit by a U.S. president to China since Trump’s own 2017 visit. The summit focused publicly on familiar pressure points in the U.S.-China relationship, including trade tensions, Taiwan, and security concerns, including the continuing conflict in Iran. But some of the most consequential dynamics shaping the relationship generated far fewer headlines, including intensifying geopolitical competition over technology, digital infrastructure, and economic security. These issues—though not explicitly mentioned in either U.S. or Chinese meeting readouts—are no less critical to understanding both countries’ strategic decisionmaking and contextualizing the summit’s true significance.

Q1: Which technology issues were (lightly) discussed at the summit?

A1: Artificial intelligence (AI) and cybersecurity, two of the most consequential technology issues in the U.S.-China technology competition, briefly surfaced in public comments, but there was little indication of meaningful progress on the core disputes between the two countries.

Heading into the summit, technology competition, specifically related to AI and semiconductor policy, was largely considered a top-priority agenda item. However, in his post-summit press call on Air Force One, the president’s comments indicated that the two sides discussed very little on the topic beyond collaboration on shared AI safety standards. U.S. Trade Representative Jamieson Greer even more pointedly said there were no discussions of chip export controls. The surprisingly muted discussion of AI and chips came despite the last-minute inclusion of Nvidia CEO Jensen Huang on the trip (he notably joined the presidential delegation in Alaska) and the current high-profile impasse over advanced Nvidia H200 chip sales to China. Interestingly, the Chinese government announced a bilateral AI dialogue as a post-summit outcome (though with no stated time frame). As of writing, the United States has issued no similar statement.

Further, while cyberattacks have long been a source of U.S.-China tension, most experts did not expect these issues to figure prominently in a summit focused on tactical trade deals. However, following the summit, President Trump made several unprecedented comments on the subject. When asked by reporters whether he raised cybersecurity with President Xi, Trump stated, “I did. He talked about attacks that we did in China. You know, what they do, we do too.” This surprising acknowledgment appears to suggest that Xi turned the conversation toward U.S. offensive cyber capabilities and away from the persistent cyber threat China poses to the United States. President Trump’s comments, in turn, suggested that he conceded parity in the two countries’ cyber activities, signifying a notable public concession for a sitting president, particularly as the intelligence community consistently frames China as “the most active and persistent cyber threat actor” targeting the U.S. government and civilian infrastructure and as the China-sponsored Volt Typhoon campaign continues to be a top concern for U.S. officials.

Q2: Which important technology issues remained in the background of the presidential discussions?

A2: Despite the fanfare surrounding the summit, several of the most consequential issues in the U.S.-China technology competition remained unresolved, and, in some cases, went completely unaddressed, according to the official readouts and subsequent interviews. AI and cyber competition rank highest among the critical issues given only a light touch during the summit, while surveillance and privacy issues, as well as the technological implications of the Iran conflict, appear not to have been raised substantively.

Q3: What AI challenges remain unresolved following the summit?

A3: Perhaps the deepest unresolved tension, before and after the summit, is structural. Both the United States and China are grappling with two AI-related challenges: First, the objective of having the best commercial AI models to prevail in their strategic competition, and second, the risk that those same advanced AI models can unleash potentially unmanageable security threats. Three major U.S. AI model developers have reported suffering distillation attacks, reinforcing that AI competition is not just about developing the most sophisticated models, but also about protecting them from competitors and mitigating the risks posed by those same models. Both governments recognize the complicated nature of these challenges, but neither has so far developed comprehensive frameworks for managing those same challenges domestically, much less diplomatically in the context of the high-stakes bilateral relationship.

U.S. export controls across administrations have limited sales of advanced chips China could use to develop more powerful AI models and have attempted to address a parallel but distinct vulnerability. Restricting chips, and therefore compute capacity, limits China’s ability to independently train the most advanced frontier models. Meanwhile, distillation attacks the United States recently accused China of conducting exploit proprietary model capabilities. Weighing these competing dynamics, the U.S. technical lead in frontier model development can be measured in mere months, as opposed to years.

Opponents of controls on advanced chip exports, like Nvidia’s Huang, argue that restricting sales cuts off U.S. firms from a massive market abroad and accelerates China’s own quest for self-sufficiency and indigenous innovation. In December 2025, President Trump shocked many by signaling that Nvidia could sell H200 chips to China. The announcement, which required the Commerce Department to take steps to implement, stopped short of permitting exports of the company’s most advanced (Blackwell) chips, yet still represented a notable inflection point in export control policy across recent administrations. Since then, however, no H200 chips have reportedly been shipped to China—despite reports that the Commerce Department has issued licenses to some Chinese firms—due to Chinese Communist Party restrictions on purchases.

Q4: Which elements of the China cyber threat were unresolved at the summit?

A4: Notably, President Trump provided public comments on cyber issues following the summit. His comments, which seemed to equate Chinese and U.S. actions in cyberspace, may have added further uncertainty in a domain characterized by heightened tensions. Whether the United States can meaningfully pursue stability with China in cyberspace—including while state-affiliated cyber actors remain embedded within U.S. critical infrastructure—remains unresolved.

The May summit took place against a backdrop of many watershed developments in the U.S.-China cyber competition. In December 2025, the administration paused plans to sanction China over Salt Typhoon intrusions—which may have exposed nearly every American’s phone data to Beijing-linked actors—amid concerns from officials that sanctioning China would hurt the broader trade deal framework struck in late 2025. The administration has yet to mount a response proportional to what many lawmakers describe as one of the most sweeping cyber espionage campaigns ever to target the United States. Further, since 2023, there has been strong evidence that Chinese state-sponsored actors have successfully infiltrated and pre-positioned within U.S. critical infrastructure, including energy, water, and communications networks. While these intrusions have not caused outages, significantly, they have demonstrated China’s interest in targeting strategic critical infrastructure for disruption, including in the event of future conflict.  

At the same time, U.S. and allied officials continue to warn that the underlying cyber threat environment is worsening. In April, the Cybersecurity and Infrastructure Security Agency (CISA) released a joint advisory with the United Kingdom’s National Cyber Security Centre and partners from Australia, Canada, Germany, Japan, and others, warning that Chinese cyber actors have shifted from using “individually procured infrastructure” to operating large-scale covert networks built from common household devices. AI is also increasingly being leveraged to accelerate and scale cyber operations. In November of 2025, Anthropic revealed the first-ever documented largely autonomous AI-enabled cyber espionage campaign conducted by China-affiliated actors, which targeted 30 entities and resulted in several successful intrusions.

These known Beijing-orchestrated activities seem to be in tension with the president’s statements and the administration’s broader cyber posture. The first pillar of the Trump administration’s new cyber strategy, “Shape Adversary Behavior,” underscores the importance of deterrence through offensive cyber operations. However, despite the 2026 Annual Threat Assessment identifying China as the “most active and persistent cyber threat,” as well as evidence of deep penetration in U.S. critical infrastructure, the strategy stops short of explicitly naming China as a priority threat actor. That omission is striking given the extent to which Chinese cyber operations, including those attributed to Volt Typhoon, have reportedly penetrated U.S. systems. The existing presence of Volt Typhoon in U.S. critical infrastructure and Salt Typhoon’s vast surveillance of telecommunications networks, along with even more recent developments, demonstrate the scale and persistence of this threat. U.S. government agencies today continue to grapple with the consequences.

Q5: What surveillance and privacy concerns went undiscussed at the summit?

A5: The growing concern among U.S. policymakers over the surveillance and privacy risks posed by Chinese technologies is another issue that went entirely unaddressed during the summit. Chinese connected vehicles have increasingly been characterized not simply as consumer products, but as sophisticated intelligence collection platforms. Modern vehicles continuously generate and transmit sensitive data, including geolocation, driver behavior, biometric information, camera feeds, and wireless communications. As Chinese automakers maintain legal obligations under China’s national security and intelligence laws, U.S. officials worry that this data is directly accessible by the Chinese state.

These privacy concerns have recently materialized in new bipartisan legislation proposed by Senator Elissa Slotkin (D-MI) and Senator Bernie Moreno (R-OH), indicating that U.S. policymakers are largely concerned about these issues. Federal rules effective as of March have already established a precedent for decoupling by banning Russian and Chinese software from all connected vehicles on U.S. roads. This move serves as a test to determine whether such a hard restriction on technological sovereignty will remain, or whether the summit indicates a path toward de-escalation. Additional legislation has also been proposed, including the Protecting Military Bases from Connected Vehicles of Concern Act, integrated into the 2025 National Defense Authorization Act, creating physical boundaries for these threats to mitigate the risk of foreign adversaries collecting sensitive military operations intelligence. This trend toward data sovereignty indicates that without any radical changes in data transparency produced from the summit, the United States is prepared to fully insulate its automative, and potentially other future markets, from Chinese technological and surveillance influence to protect U.S. national security.

Q6: What questions stemming from the Iran conflict went undiscussed at the summit?

A6: The war with Iran, launched on February 28, is quickly approaching its three-month mark, and there remains significant ambiguity surrounding whether the United States and Israel have achieved their stated military and diplomatic objectives. What has been made clear, though, is that as the war continues, China is watching closely and learning its own lessons from the conflict.

Chinese-linked firms have reportedly leveraged AI and open-source intelligence to sell information about U.S. force movements and military activity. These activities reflect a broader trend within China’s civil-military integration strategy, under which firms have been largely supported by China to advance its goal of developing AI with defense applications. They also speak to the growing competition in the space domain, as China-operated low-earth orbit satellites are proliferating, despite the longstanding U.S. lead in these technologies.

Given China’s close attention to U.S. military tactics throughout the conflict, another important consideration is the lessons that Beijing may draw from the role of U.S. digital capabilities in modern warfare, as exhibited by the conflict. U.S. military leaders repeatedly emphasized the role of cyber and space capabilities as key enablers for air operations in the early days of the conflict, underscoring the importance of these capabilities in preparing the operational environment, enabling coordination, and supporting precision targeting.

From Beijing’s perspective, these capabilities likely reinforce the strategic importance of targeting digital infrastructure, including data centers and space-enabled systems, in this and any future conflicts. While Iran’s capabilities in these domains remain comparatively muted and it has demonstrated a limited capacity to carry out destructive attacks, China poses a significant threat through its advanced cyber, electronic warfare, and counterspace abilities. The conflict therefore offers Beijing an opportunity to study how the United States and partners leveraged cyber, space, and information operations in the early stages of the conflict. It also likely provides opportunities to understand U.S. interdependencies and reliance on digital and physical infrastructure, and where those interconnections might be vulnerable to disruption.

In this vein, the United States would do well to monitor the apparent role of the Chinese BeiDou satellite navigation network in supporting improvements in Iranian missile targeting and navigation capabilities during the conflict. These reports emphasize China’s prioritization on decoupling from the U.S.-operated GPS constellation, which is particularly vulnerable to electromagnetic warfare attacks.

Q7: What did the summit reveal about the trajectory of technology competition between the United States and China? What can we expect from the upcoming September summit?

A7: The relatively light treatment of critical technology issues during the Trump-Xi summit stands in stark contrast to these issues’ importance in the two nations’ geopolitical competition. It also emphasizes how challenging—and inconducive to symbolic, leader-level engagements focused on optics over substance—technology policy issues can be to unstick. Most notably, there was little movement on rapid-paced AI model developments and chip export controls in the meetings, despite Jensen Huang’s highly publicized presence and these issues’ outsized importance to the Trump administration’s domestic and foreign policy agendas. On the other hand, President Trump’s comments suggesting parity in U.S. and Chinese offensive cyber operations were a surprising concession, given China’s well-documented and persistent campaigns targeting U.S. government and civilian critical infrastructure over many years.

Looking ahead to the announced September meeting between the two presidents, these tensions—as well as surveillance and privacy concerns, strategic learnings from the ongoing conflict in the Middle East, and more—will figure prominently in the background. None of these issues lend themselves to quick fixes. However, the next summit can serve as a helpful barometer for whether both nations can tackle short-term trade challenges at the same time as respective delegations attempt to make progress on some of the most important, and systemic, technology challenges facing the world today. For the United States, getting ahead of technology challenges from China must begin with robust interagency processes designed to vet and put forward clear national policy positions on AI leadership and export controls, cyber and digital domain threats, as well as digital sovereignty and surveillance.

Lauryn Williams is the deputy director and senior fellow in the Strategic Technologies Program at the Center for Strategic and International Studies (CSIS) in Washington, D.C. Kuhu Badgi is the program coordinator and research assistant for the Strategic Technologies Program at CSIS.