Trump’s China rivalry and “AI race” delusion may endanger US, experts say
Ahead of a meeting between Donald Trump and Xi Jinping, US officials said Washington and Beijing have discussed an AI safety notification mechanism to allow either side to alert the other about dangerous or unpredictable AI behavior. Experts and lawmakers expressed skepticism, noting the proposal lacks a technical review body and may be undermined by broader US-China tensions over export controls and recent legal accusations against Chinese AI firms.
Why It Matters
Cooperation between the two leading AI powers could help manage shared risks from advanced systems, but mutual mistrust and limited scope in proposed arrangements may prevent meaningful information-sharing on urgent safety issues. How Washington and Beijing structure any notification channel will influence international AI governance efforts.
Key Facts
- Leaders meeting: Donald Trump and Xi Jinping scheduled to meet Thursday and Friday
- Announcement: Treasury Secretary Scott Bessent said both sides discussed creating an AI safety notification mechanism
- Purpose of mechanism: A channel for either country to send alerts when their AI systems pose threats or behave unpredictably
- US actions increasing tension: Trump administration has expanded export controls and Congress is considering additional chip/hardware restrictions
- DOJ accusation: U.S. Justice Department recently accused six Chinese AI firms of stealing from U.S. frontier labs
U.S. and Chinese officials have reportedly discussed setting up an AI safety notification mechanism ahead of a bilateral meeting between Donald Trump and Xi Jinping. The proposed channel would allow either side to flag instances where AI systems in their country present hazards or act unpredictably, and officials said it could also facilitate sharing a "vision of common goals." Treasury Secretary Scott Bessent announced the discussions as part of preparatory diplomacy for the leaders' talks.
Analysts and lawmakers expressed reservations about the plan's current form. Sources familiar with Bessent’s proposal told Politico that the mechanism, as described, contains no standing technical body to evaluate the urgency or validity of incoming alerts. Sam Bresnick, a research fellow at the Center for Security and Emerging Technology, said the absence of expert interlocutors would limit the channel’s usefulness, particularly for assessing military or high-risk technical issues.
The initiative comes amid deepening geopolitical friction between Washington and Beijing. The Trump administration has tightened export controls aimed at limiting China’s access to technology relevant for AI, and U.S. lawmakers are considering additional restrictions on chips and hardware. Adding to the strain, the Justice Department recently accused six Chinese AI companies of stealing from American frontier research labs, an allegation China has disputed. Observers warned these broader tensions could make Chinese counterparts reluctant to share sensitive safety information.
Some experts described the notification channel as a modest but helpful first step if it were expanded. Su Lian Jye of Omdia said alerts could promote transparency but would not erase concerns over issues like technology transfer, pricing competition, and research pacing. Alvin Wang Graylin, who has advised on AI policy, called it a positive development that both sides scheduled follow-up talks in November, while also noting that meaningful cooperation would require agreement on risk classifications, response protocols, and technical review processes. Critics including Representative Ro Khanna urged that talks go beyond a basic line of communication to include stronger international commitments on high-risk AI capabilities.
Keep Reading
If Amazon doesn’t know how to eliminate carbon emissions, then who does?
Inflation could cost Republicans the Senate — this chart shows how
This test screens for 50 cancers — but whether Medicare and other health insurers will cover it hinges on FDA approval
