The 13th round of the China–U.S. Dialogue on Artificial Intelligence and International Security was successfully held in Dubai on November 15–16, 2025. CISS Director Da Wei and former U.S. Deputy Secretary of Defense Colin Kahl attended the event and delivered remarks. More than 30 members from both sides participated, engaging in in-depth exchanges on issues related to AI and international security.

This round of dialogue focused on an AI risk-tiering framework. Experts from both countries discussed risk-management measures based on specific scenario cases, enhancing mutual understanding of concepts and terminology related to “loss of control.” Participants also conducted scenario exercises on two issues: the threats posed by the rise of non-state actors to national security and inconsistencies in AI system behavior. In addition, building on the consensus reached at the China–U.S. presidential meeting in Lima, both sides explored topics concerning the intersection of AI and the nuclear domain.

Moving forward, the project teams will continue to build on the outcomes of this round, delving deeper into forward-looking issues in AI safety. Both sides aim to continuously expand mutual understanding and consensus in the AI field and jointly address the potential security risks and challenges posed by emerging AI technologies.