The Pulse
Bessent Discusses AI Incident Channel After U.S.-China Summit Deal
Treasury Secretary Scott Bessent described a proposed U.S.-China notification process for AI incidents in an interview published October 3. The two governments had already agreed at the September summit to establish a bilateral channel and
AI.info Team · Written with AI ·
“I think that they didn't realize how powerful their open source models are.”
Scott Bessent, U.S. Treasury secretary
Bessent made the remark about Chinese AI models in an interview with The Axios Show published October 3, while discussing a process for Washington and Beijing to notify each other if an AI incident becomes serious. He said he believes China would agree to such a process. But the two governments had already publicly agreed to establish a bilateral AI-incident channel during President Donald Trump and President Xi Jinping’s September state visit.
A proposal that followed a summit agreement
Axios reported that Bessent and Chinese Vice Premier He Lifeng discussed AI safety before Xi’s visit, including a proposed communication channel. In the interview, Bessent described the goal as a notification process for when something goes wrong with AI. The September summit statements went further: the White House said both countries agreed to establish a bilateral communication channel for incidents involving what it called “super intelligence.”
China’s foreign ministry listed the same agreement among eight outcomes of Xi’s September 23–25 visit: the two sides agreed to establish a bilateral channel for AI incidents. The White House also said the countries had established a U.S.-China Super Intelligence Dialogue to discuss risks and benefits. Its next exchange is scheduled for November 2026.
The agreement sets a channel, not its operating rules
The public announcements establish that both governments agreed to create a channel, but they do not set out how it will work. They do not specify which incidents would trigger a notification, which officials or agencies would send one, or how quickly either side would be expected to respond. Those details matter because a system that covers only events with national-security consequences would have a narrower remit than a general warning channel for serious AI failures.
Bessent’s October comments describe the proposal in broad terms, rather than supplying those operating rules. He also said the United States should put more attention on defending critical systems from AI threats, not only on building more powerful models. Asked about AI executives warning they could lose control of advanced systems, Bessent replied: “Well, then they should slow down.”
Beijing and Washington frame AI cooperation differently
The White House announcement uses “super intelligence” in describing both the new dialogue and the incident channel. China’s foreign ministry said the two leaders discussed AI in depth and that Beijing respects Washington’s choice of terminology, while calling for further exchanges and discussion. That leaves room for cooperation on communications without suggesting the two governments share a single account of AI risks.
The immediate test is whether the agreed channel gains a defined scope and a working process alongside the broader dialogue. For now, the public timetable contains one fixed date: the next U.S.-China AI exchange is due by November 2026.