An international summit is a complex affair. Before Presidents Donald Trump and Xi Jinping meet later this week, American and Chinese officials hold early discussions about the details. The talks-before-the-talks on Sunday featured U.S. Treasury Secretary Scott Bessent and Chinese Vice Premier He Lifeng, and Bessent floated the idea of a notification system for early warning of AI incidents.
Reports are vague on the details, and we mostly just have Bessent’s word to go on. The New York Times quotes Bessent saying that the proposed channel would cover incidents at “the national security level,” and he elaborates in CNBC: “so both sides can agree on what the leading AI dangers are, whether it’s uncontrollable agents, whether it’s non-state actors in cyber, non-state actors in bio weapons.”
Although these AI dangers compete with trade and tariffs on the summit agenda, it is good to see them considered, and a rapid incident-reporting system seems a promising early step if implemented well. It echoes Haven’s guest post yesterday, which recommended “establishing a secure U.S.-China incident channel,” and the precedent set by such a channel could lay groundwork for further much-needed bilateral action. Bessent reportedly also said that he and He Lifeng would likely meet again in Shenzhen, China, in November.
The analyses and opinions expressed on AI StopWatch reflect the views of the individual contributors and the sources they cover, and should not be taken as official positions of the Machine Intelligence Research Institute.