On February 13–14, 2025, the Center for International Security and Strategy (CISS) at Tsinghua University and the Brookings Institution jointly hosted U.S.-China Track II Dialogue Round XII on Artificial Intelligence and International Security in Munich, Germany. Former Vice Minister of Foreign Affairs Fu Ying and Brookings Institution President Cecilia Rouse attended and delivered speeches. Nearly 40 members from the Chinese and American project teams participated in discussions on topics related to artificial intelligence and international security.

This round of dialogue was divided into three discussion sessions.The first session focused on the terminology related to artificial intelligence risks and was co-chaired by Xiao Qian, Deputy Director of CISS, and Tucker Hamilton, former Chief of AI Test and Operations for the U.S. Air Force. Experts from both sides exchanged views on over 20 AI-related terms, including catastrophic risks and existential risks, aiming to refine risk classification for a better understanding of the potential security threats posed by AI technology.

 The second session centered on the stratification of AI risks and was chaired by Michael Sellitto, Director of Geopolitics and Security Policy at Anthropic. Participants from both expert groups shared their perspectives on AI risk stratification, discussed numerous AI risk scenarios, and explored ways to establish a more comprehensive risk assessment framework.

The third session featured an AI scenario simulation exercise, chaired by Tom Stefanick, Visiting Scholar at the Brookings Institution. Experts from both sides examined the application of AI technology in command and decision-making systems and discussed measures to build trust and mitigate risks.

Additionally, a dinner discussion session was held, where Xue Lan, Dean of the Institute for AI International Governance at Tsinghua University and a CISS Academic Committee member, introduced the development of China's AI sector and answered questions from participants.

The dialogue concluded with summary remarks from Andrew Forrest, Founder of the Minderoo Foundation; Chen Qi, Deputy Director of CISS; and Ryan Hass, Senior Fellow at the John L. Thornton China Center at the Brookings Institution. They noted that this round of discussions deepened mutual understanding of AI-related risk concepts and enhanced both sides’ awareness of risk prevention and response. They expressed hope that project team members would continue advancing discussions along existing topics, producing further research outcomes, and jointly addressing the potential risks and challenges posed by AI.