U.S. and China set up AI security dialogue ahead of summit

U.S. Treasury Secretary Scott Bessent and Chinese Vice Premier He Lifeng agreed to create a U.S.-China AI dialogue to handle security incidents, officials said, days before a planned Trump‑Xi summit this week. Markets were watching the diplomatic move as investors count down to the meetings; reaction pending.

U.S.-China AI security channel

Officials said the channel is intended to provide a direct line for discussing AI-related security incidents and reducing the risk of miscalculation. The announcement, they added, followed bilateral talks in which both sides described a need for clearer procedures to manage emerging AI risks.

The setup stops short of joint norms or export controls; officials framed it as a crisis-management mechanism rather than a regulatory pact. That narrows immediate expectations for trade or investment policy changes while creating a structured route for incident reporting and deconfliction between Washington and Beijing.

Watchers said the channel could lower the odds of an accidental escalation if it is used promptly and transparently; its effect will hinge on staffing, information-sharing rules and political buy-in on both sides. Officials said further details would be worked out in follow-up meetings.

Expect statements and working-level exchanges during the summit week, Sep. 21–27, 2026, when leaders meet in Washington and officials are due to elaborate on the channel's scope and operating rules.