US Proposes AI Incident Alert System to China
Bilateral talks aim for transparency between top AI powers on security threats.
1Proposal for AI transparency
US Treasury Secretary Scott Bessent proposed a bilateral AI incident notification system during talks with Chinese Vice Premier He Lifeng in Manhattan. The mechanism would cover AI-related national security events but excludes semiconductor export controls—a key Chinese demand. This marks the first concrete step toward operational cooperation between the world's top two AI powers.
2Strategic meeting logistics
Discussions occurred at JPMorgan's headquarters after UN General Assembly security restrictions limited midtown venue options. While the bank provided space, it wasn't involved in substantive talks. Working-level teams continued negotiations into Monday, focusing on AI governance frameworks ahead of Xi Jinping's scheduled White House visit later in the week.
3China's guarded response
Chinese trade negotiator Li Chenggang described talks as 'not bad' but Beijing hasn't accepted the US proposal. Xinhua's readout mentioned 'candid' AI discussions without addressing the notification system. Both sides agreed to future AI talks but set no details—reflecting China's pattern of delaying binding commitments on sensitive technologies.
4Parallel trade developments
The US-China Board of Trade, conceived during Trump's 2025 Beijing visit, became operational during these talks. It will negotiate tariff reductions on non-strategic goods, though specific products and rates remain undisclosed. This creates a potential linkage where AI cooperation could influence trade concessions.
5Imminent deadline pressures
Two critical deadlines loom on November 10: expiration of the US-China tariff truce and China's rare-earth export controls suspension. American manufacturers report insufficient magnet shipments despite the controls being lifted last October, adding urgency to these negotiations.
6US domestic policy context
Bessent's proposal follows his congressional testimony advocating creator liability for AI systems, opposing liability shields sought by US labs. FTC Chair Andrew Ferguson recently warned against regulatory trade-offs for antitrust relief—signaling tough US oversight that could shape future AI agreements with China.
7Upcoming diplomatic moves
These talks set groundwork for Xi Jinping's White House visit, potentially accompanied by BYD, CATL and Xiaomi executives. The AI proposal's timing suggests the US wants deliverables before the summit. Beijing hadn't formally confirmed Xi's visit as of last week, maintaining negotiating leverage.
8Strategic implications
The proposed AI alert system represents a test case for US-China tech cooperation amid fierce competition. Its exclusion of semiconductor controls shows Washington's red lines, while China's noncommittal response reflects distrust. Success would require overcoming fundamental differences in AI governance approaches and verification mechanisms.
Original article: The Next Web — US Proposes AI Incident Alert System to China