In the lead-up to Chinese President Xi Jinping’s visit to Washington for a summit with U.S. President Donald Trump, officials from both countries made a notable advance by agreeing to establish a formal dialogue on artificial intelligence (AI) security. This development, announced shortly before Xi’s arrival, followed talks between U.S. Treasury Secretary Scott Bessent and China’s Vice Premier He Lifeng. Alongside instituting a regular communication channel, the United States proposed a mutual notification mechanism for AI incidents posing serious national security risks—echoing the Cold War–era nuclear hotline between Washington and Moscow.
These steps represent tentative efforts to introduce safeguards amid accelerating technological competition between the two powers, which are vying for leadership in AI development. Although concerns about AI turning against humanity have been widely publicized, the more immediate challenge lies in managing strategic rivalry without triggering destabilizing risks.
President Trump has publicly dismissed doomsday predictions regarding AI as a “hoax,” while Beijing perceives U.S. calls to limit advanced AI models as attempts to hinder China’s progress and confine it to a secondary role. However, Chinese officials have also sought to reassure the global community about AI safety, given that Chinese open-source AI technologies have become significant exports. U.S. officials contend that Beijing’s primary worry is less about the existential risks of AI and more about the technology’s potential impact on the Chinese Communist Party’s domestic control.
Despite persistent distrust, both governments recognize shared vulnerabilities that necessitate cooperation. A key concern is the risk that highly capable AI technologies could be acquired or misused by non-state actors, including terrorists, rogue groups, or organized crime networks. Such AI capabilities might facilitate activities ranging from bioweapons development to cyberattacks on critical infrastructure and financial systems. There is also the danger of AI systems escaping controlled environments and autonomously conducting cyber intrusions.
Experts have advised establishing baseline standards to oversee AI access, verify users, rigorously test models, and implement containment protocols to mitigate these risks. Given the rapid pace at which AI-driven incidents could escalate—potentially compressing response times and clouding attribution—U.S. officials see the proposed AI incident notification pact as a critical measure to avoid unintended escalation.
The summit agenda also includes discussions about AI integration into weaponry and nuclear command systems. In a related development from November 2024, President Xi and then-President Joe Biden agreed that humans should retain ultimate control over nuclear weapons use. However, no significant progress on formalizing such principles has occurred since, even as AI technologies have advanced rapidly. Security specialists from both nations have drafted recommendations since 2019, advocating for “absolute” prohibitions on autonomous AI decisions to launch nuclear arms and mandatory human authorization for AI-driven cyberattacks on nuclear command networks.
Observers note that the establishment of a U.S.-China AI dialogue and a communication hotline could set a precedent analogous to the Cold War nuclear hotline—a mechanism designed to manage crises before they escalated to conflict.
Beyond the technical and security dimensions, the summit carries symbolic weight as a test of mutual trust between two geopolitical rivals. Analysts point out that China’s internal diversity—in which some regional officials prioritize pragmatic economic engagement while others emphasize nationalistic competition with the United States—adds complexity to negotiations. Moreover, both Washington and Beijing harbor suspicions about each other’s intentions, shaped by longstanding insecurities and historical experiences.
While some in the U.S. advocate for decoupling from China to curb its technological rise, experts warn that such policies could provoke China to accelerate self-reliance, particularly in critical sectors like rare earth elements and semiconductors. The need to engage China realistically, acknowledging its unique civilizational identity and internal challenges, is emphasized as essential to avoiding renewed confrontation.
Ultimately, the AI dialogue and the broader summit represent an opportunity for the two leaders to choose collaboration over suspicion. Their ability to establish foundational agreements on AI security could help stave off an accelerated slide into a cold war–style rivalry, providing a framework for managing fast-evolving technologies whose misuse might have far-reaching consequences. The summit thus embodies a critical moment in shaping whether U.S.-China relations evolve toward cooperation or deepen into confrontation.