US proposes AI incident alerts in talks with China

U.S. Treasury Secretary Scott Bessent said the United States proposed a mechanism for the United States and China to notify each other about artificial intelligence incidents that reach the level of a national-security threat. The proposal came during bilateral talks in New York on September 20. No operating arrangement, reporting protocol or Chinese commitment has been announced.

China’s official account of the talks confirmed that the two sides discussed AI-related issues, but did not mention the notification mechanism, its terms or Chinese acceptance. For now, the public record shows an exploratory governance proposal rather than an agreed system.

Bessent joined U.S. Trade Representative Jamieson Greer and Chinese Vice Premier He Lifeng in the talks. Bessent described the proposal as part of a broader U.S.-China AI dialogue and said representatives would meet again to discuss common goals and threats.

Several basic questions remain unanswered. The available accounts do not say which incidents would trigger notice, which agencies would communicate, how quickly notification would be required, what response would follow or whether an arrangement would be legally binding. They also do not establish a treaty, memorandum, formal channel, enforcement mechanism or launch date.

The AI discussion was part of wider economic and trade consultations. Bessent said U.S. export controls were not discussed in the weekend meetings. Greer said controls on advanced-chip exports to China and semiconductor-manufacturing equipment were outside the proposed notification mechanism.