US and China to resume AI safety talks amid mounting rivalry
After a two-year pause, the United States and China plan to restart dialogue on artificial-intelligence safety, aiming for mutual visibility rather than a binding pact.
U.S. and Chinese officials are set to reconvene AI safety negotiations after a two-year hiatus, with a meeting likely to coincide with the anticipated Trump-Xi summit. The objective is not to forge a comprehensive agreement but to provide each side enough insight into the other’s AI risk assessments and mitigation strategies, preventing unchecked acceleration. Beijing’s state-linked account Yuyuan Tantian cautioned that such talks might become a tool to restrain China’s technological rise, echoing past U.S. national-security justifications for export controls.
Analysts note that mutual visibility, rather than trust, can curb the worst-case race dynamics, and that informal Track 1.5 and 2 dialogues have already taken place. Recent statements from OpenAI and Anthropic call for voluntary slowdowns, underscoring the urgency for shared safety standards. The success of the talks will hinge on keeping AI risk mitigation insulated from broader trade and security negotiations.
Why it matters
AI safety talks could shape global norms that prevent a risky AI arms race between the world's two biggest tech powers.
In this story
