US and China Set November AI Safety Talks in Shenzhen
The two sides agreed to a formalized dialogue and an incident line for AI safety emergencies, Treasury Secretary Scott Bessent said after Sunday talks in Washington.
WASHINGTON, Sept 21 (Reuters) - Senior US and Chinese officials will meet again in about two months in Shenzhen, China, to discuss artificial intelligence dangers and communications protocols in the event of AI safety incidents, US Treasury Secretary Scott Bessent said on Monday.
Bessent told CNBC that during talks on Sunday with Chinese Vice Premier He Lifeng, the two sides agreed to set up a formalized dialogue on AI, including an "incident line" for communications about AI safety incidents.
"And then we want to start discussing protocols ... so both sides can agree on what the leading AI dangers are, whether it's uncontrollable agents, whether it's non-state actors, and cyber non-state actors," Bessent said.
Bessent reiterated his view that AI developers need to be held liable for their creations, including situations in which AI agents go rogue and stage uncontrolled hacking incidents like the attack on open-source AI tool platform Hugging Face.
"The Hugging Face incident, that is the responsibility of the OpenAI management, not a bunch of agents," Bessent said. "These labs need to take responsibility for themselves. They can slow down anytime they want to."
Asked about the future of the US-China trade truce that capped bilateral tariffs but which is set to expire on November 10, Bessent said the two sides discussed it but did not specifically say that they had agreed to extend the deadline.
Bessent had positive comments about the US-China trade relationship.
"That was a focal point, and we again do take great stability in the relationship since last fall, and I think we're going to maintain that," Bessent said.
(Reporting by David Lawder; Editing by Louise Heavens and Will Dunham)
