Promising Topics for US–China Dialogues on AI Safety and Governance

20 January 2025

Saad Siddiqui, Kristy Loke, Stephen Clare, Marianne Lu, Aris Richardson, Lujain Ibrahim, Conor McGlynn, Jeffrey Ding

View Policy Paper

In 2023, both the US and China signed the Bletchley Declaration, acknowledging the potential for serious harm from advanced AI systems as well as the importance of cooperating to mitigate it. Since then, the US and China have also jointly signed UN resolutions on AI issues, held a round of intergovernmental dialogues on AI, and agreed to limit the integration of AI into control systems for nuclear weapons. The number of Track II dialogues has similarly increased. As these initiatives continue and as AI capabilities continue to advance, it is important to ask: what topics would give future dialogues the best chance of making meaningful progress on important AI governance issues?

In this report, we develop recommendations for AI governance and safety dialogue topics from one specific angle: identifying topics on which there is significant common ground between the US and China, through a comprehensive analysis of over 40 key primary AI policy and corporate governance documents from the two countries. We analyze these areas of common ground in the context of US–China relations, setting aside topics that would be deemed too sensitive, or linked to existing tensions between the two countries.

Through this analysis, we arrive at recommendations for dialogue topics and suggestions for which diplomatic tracks (i.e. Track I or Track II) could adopt these recommendations. While the focus of our recommendations is the bilateral context, many of these discussions could also take place in the multilateral context, through fora such as the AI summit series.

Our four key recommendations are:

  1. US and Chinese governments should strengthen existing intergovernmental dialogue, covering issues related to national security, such as evaluating models for ‘dangerous capabilities’ and preventing proliferation to non-state actors.
  2. The US and Chinese governments and their respective companies should explore technical standards-setting discussions on reliability, robustness and adversarial testing, either in existing standards-setting bodies (e.g. the International Standards Organisation) or new fora.
  3. International industry consortia should consider either involving Chinese companies in existing dialogues or setting up distinct tracks of dialogue with Chinese actors, particularly on content provenance.
  4. Track II dialogues should include discussions on emerging or novel approaches to safety and governance, especially those for which there may be existing common ground between US and Chinese perspectives.