U.S. Treasury Secretary Scott Bessent says he plans to propose an AI incident notification process with China, but no agreement or working channel has been announced. His October 3 remarks to Axios put the idea back in the spotlight, not into effect.

Bessent said he had discussed an AI incident channel with Chinese Vice Premier He Lifeng ahead of President Xi Jinping’s state visit. He said he believes Beijing may agree. That is Bessent’s assessment, not a commitment from China, according to Axios.

A proposal, not a working hotline

The idea dates back at least to September. On September 20, Bessent said Washington had proposed a notification mechanism for AI incidents that could affect national security during talks with He in New York. The Associated Press reported the proposal on September 21.

Key details remain unanswered. No public rules explain which incidents would trigger a warning, who would send it, or how either government would respond. The two sides have announced no agreed process or operating details.

What a channel could change

If both governments accept and build the channel, it could give officials a way to alert each other about serious AI-related incidents with national-security consequences. The proposal described in September has a limited scope: it would not set shared AI safety standards or resolve wider disputes over AI development.

For now, the concrete development is diplomatic. Bessent is publicly backing the idea again, while China’s position and the practical terms remain unannounced.

The proposal could create a way to communicate about serious AI incidents, but it is not an agreement, a working hotline, or a set of shared safety rules.