US proposes AI incident alert system with China

The US has proposed a “notification mechanism” for AI incidents that could affect national security, Treasury Secretary Scott Bessent said.

Bessent raised the idea after meeting Chinese Vice Premier He Lifeng in New York on Sunday. The two sides also discussed a broader US–China dialogue on AI. The notification proposal would address a different need: alerting the other government when a serious incident occurs.

Tsinghua University researcher Sun Chenghao told the Associated Press that shared risks could include AI-enabled cyberattacks, biological misuse and major model failures. “It would be a mistake to dismiss the entire safety debate as geopolitical competition,” he said. Sun also noted China’s growing attention to the safety of AI agents and the risk of losing control over AI systems.

Presidents Donald Trump and Xi Jinping discussed possible AI consultations in May, but no formal forum followed. They are expected to consider the US proposal when they meet in Washington this week.