US-China Agree on AI Safety Dialogue After Eight Hours of Talks

The US and China have agreed to establish a formal channel for warning each other about AI incidents that could affect national security, in a first-of-its-kind arrangement. The agreement was reached after eight hours of talks at JPMorgan Chase's Manhattan headquarters between Treasury Secretary Scott Bessent and Chinese Vice Premier He Lifeng.

The proposed mechanism would create a notification system allowing each government to alert the other when an AI-related incident rises to the level of national security concern, covering everything from autonomous weapons behavior to infrastructure penetration. However, export controls on advanced semiconductors and AI chip manufacturing equipment remain outside the arrangement's scope.

The talks also touched on trade issues, including China's underdelivery on rare-earth export commitments made under the Busan trade truce. The US is focused on Chinese purchases of US energy, agricultural goods, and potentially medical devices, while Beijing is protecting high-value exports.