US Treasury Secretary Scott Bessent proposed a new communication mechanism to Chinese Vice Premier He Lifeng during their meeting in New York on September 20. The initiative aims to establish a direct channel for sharing information regarding artificial intelligence incidents that could threaten national security.
The proposed framework focuses on defining dangerous scenarios and creating operational protocols. Key areas of concern include autonomous AI agents, the use of AI systems by non-state actors, and cyberattacks. Currently, no formal protocol exists between the two nations for managing such risks.
Following the discussions, both parties agreed to continue official dialogue on AI safety. The next meeting is scheduled to take place in Shenzhen in two months, although the specific date and format remain undecided. Beijing has not yet confirmed its acceptance of the proposal.
While the talks covered AI as part of broader economic and trade discussions, the US clarified that the negotiations did not include export controls on AI chips or semiconductor manufacturing equipment. The proposed incident channel remains a subject of ongoing diplomatic negotiation.
