The US Proposes AI Incident Alert System to China
The United States (see details) has proposed that Washington and Beijing notify each other of AI incidents serious enough to qualify as national security events. This idea was presented by Treasury Secretary Scott Bessent to Chinese officials during talks in Manhattan on Sunday.
Bessent, along with Jamieson Greer, the trade representative, met with He Lifeng, the Vice Premier, and Li Chenggang, the trade negotiator, at JPMorgan's headquarters. The discussions continued into Monday.
According to Bessent, moving "from opaque to more transparency between the No. 1 and the No. 2 AI powers in the world is very important." He emphasized it as a channel for communication rather than a treaty. Greer clarified that export controls on AI chips and semiconductor manufacturing equipment, a key demand from Beijing, were not included in the proposal.
Nikkei reported that the US side intends to address AI-directed cyberattacks and misuse of agentic systems, requesting laboratories on both sides to share threat information. Li described the talks as "not bad" and confirmed AI was among the discussed topics, along with trade and investment.
While both sides agreed to further meetings regarding AI, no date, location, or format has been set. On the trade front, Greer stated that the Board of Trade, agreed upon during Trump's May visit to Beijing, "has been made operational." It aims to negotiate tariff cuts on non-strategic goods in energy and agriculture.
The proposal comes after Bessent's statement to the House Financial Services committee, where he rejected liability shields for US laboratories and emphasized that "the creators are liable for what they build and generate" as the best guarantee of safety.
These discussions set the stage for President Xi Jinping's upcoming visit to the White House on Thursday, during which executives from BYD, CATL, and Xiaomi may accompany him.