US and China Strategic AI Dialogue (SAID) Proposal

The US and China have put forward a proposal to establish a permanent, high-level intergovernmental mechanism on AI safety and governance called the Strategic AI Dialogue (SAID). The dialogue will be led by Treasury Secretary Scott Bessent on the US side and Vice Premier He Lifeng on the Chinese side; White House Office of Science and Technology Policy (OSTP) Director Michael Kratsios will also participate continuously to coordinate technical work. Quarterly meetings will take place in a neutral country, preferably Switzerland, with working groups operating continuously in between.

The proposal aims to reduce AI risks affecting both countries that cannot be prevented unilaterally (nuclear command and control, biological catastrophes, advanced model safety, financial stability, employment disruptions, cyber risks, and AI-enabled crimes) through strategic communication, technical exchange, crisis notification, and common terminology development. SAID does not aim to resolve technological competition but focuses on risk reduction by addressing mutual security vulnerabilities; it builds on the first intergovernmental AI meeting in Geneva in May 2024 and the presidential-level nuclear control commitment in November 2024.

The structure centers high political leadership to avoid the bureaucratic gridlock of previous dialogues; the Bessent-He duo provides authority to bring together economic, financial, security, and technological agencies. OSTP leads the technical work on advanced model assessment, biotechnology, standards, and scientific research security. SAID is designed as a narrow, functional risk reduction mechanism to prevent shared catastrophes, rather than a comprehensive AI governance negotiation.