On September 29, the Center for International Security and Strategy (CISS) of Tsinghua University, together with the Brookings Institution and the Minderoo Foundation, hosted the 8th round of “US-China Dialogue on Artificial Intelligence (AI) and International Security” in Milan, Italy. Colin Kahl, former Under Secretary of Defense for Policy, gave a speech for the Dialogue. More than 30 members of the project teams from China and the U.S. and guests participated in the Dialogue and had an in-depth discussion on topics related to AI and international security.
The Dialogue included three sessions. The first session focused on the discussion of core conceptual terminology related to AI, which was moderated by Xiao Qian, Deputy Director of CISS. Participating experts conducted in-depth discussions on the core terminology provided by both sides and agreed to continue consultations in the joint working group on a regular basis to promote exchanges in AI-related terminology and concepts between the two sides and narrow the differences in mutual understanding.
During the second session moderated by Michael Sellitto, Head of Geopolitics and Security at Anthropic, the Dialogue focused on AI military application scenarios. Taking into account specific cases, the participating experts explored scenarios such as heterogeneous UAV swarms and natural disasters and large-scale language modeling. They analyzed the measures to reduce risks and avoid algorithmic dependence under different AI military application scenarios. Experts from both sides agreed to continue to strengthen group discussions to develop common scenarios and further analyze the possibility of reducing AI risks in specific scenarios.
The Dialogue was concluded by Chris Meserole, Director of the Artificial Intelligence and Emerging Technology Initiative at the Brookings Institution, and Zhu Qichao, Deputy Director and Fellow of the Institute for Defense Technology and Strategic Studies at the National University of Defense Technology. The two sides agreed to strengthen communication in the area of AI security governance, explore ways to seek common ground while reserving differences at the international governance level, and continue to promote the research of the Terminology Working Group. At the same time, the two sides will also clarify the consensus and differences of the expert groups on both sides based on the results of the past eight rounds of dialogues and release a report on the joint research results in due course.
In October 2019, CISS and the Brookings Institution jointly established the “US-China Dialogue on Artificial Intelligence and International Security” project, with eight rounds of dialogues having been organized so far. The two sides have organized experts to conduct continuous discussions on topics related to AI and international security, enhancing communication between Chinese and US experts, promoting both sides to resolve differences, facilitating understanding and cooperation, and exploring feasible paths to jointly address the risks and challenges of AI security.