US, China to meet again on AI safety in two months in Shenzhen, Bessent says - Reuters

US Treasury Secretary Scott Bessent said so on Monday, according to Reuters. Bessent also said the two sides had agreed to set up a formalised dialogue on AI.

That dialogue is to include an 'incident line', which Bessent described as a channel for communications about AI safety incidents. The term 'incident line' is the phrase used in the report for the mechanism; its design, participants and operating rules are not described. The next meeting is given a location — Shenzhen — and a timeframe of roughly two months, rather than a fixed date. The subject matter listed for that meeting is AI dangers and the protocols the two governments would follow when communicating about AI safety incidents.

The article presents the Treasury Secretary as the source of the account and Reuters as the outlet carrying it. No further details were given in the report on the membership of the delegations, the agenda, or the date on which the Shenzhen meeting would be held.