U.S. Treasury Secretary Scott Bessent proposed an artificial-intelligence incident notification mechanism during talks in New York with Chinese Vice Premier He Lifeng. The concept is intended to improve transparency when an AI-related event could create national-security risks.
Details remain limited, including what kinds of incidents would trigger a notice, how quickly information would be shared, and whether independent verification would be required. The proposal nevertheless represents an effort to build a basic safety channel between the world’s two leading AI powers.
Technology talks join the broader economic agenda
The discussions also covered trade and preparations for a planned meeting between Presidents Donald Trump and Xi Jinping. Officials announced a new framework to carry forward negotiations on tariffs and commerce.
An effective AI alert system would require trust at a time of strategic competition over chips, models, cybersecurity, and military technology. Even a narrow agreement could provide a template for managing unexpected technical failures or dangerous misuse.