Nations won't agree to AI governance restrictions if they think their rivals will cheat on the agreement. Besides verification mechanisms (enacted during and after the adoption of the agreement), what confidence-building measures can pave the way for nations that distrust each other to successfully collaborate on AI R&D regulations?
About the project
MIRI has proposed an international agreement to halt the development of artificial superintelligence until it can be done safely. For details, see the proposal here: https://arxiv.org/abs/2511.10783
The agreement requires each nation to restrict AI development within its own borders, but nations won’t agree to do this if they think other countries will cheat on the agreement. Besides verification mechanisms (enacted during and after the adoption of the agreement), what confidence-building measures can pave the way for nations that distrust each other to successfully collaborate on the AI R&D halt?
This project will explore how nations and AI labs can establish the trust and transparency needed to forge and uphold international agreements that reduce existential AI risks. We will investigate mechanisms for confidence-building – drawing lessons from historical arms control, WMD treaties, and climate agreements – and apply them to the AI context. The researcher will analyze why some international agreements (e.g. nuclear and chemical arms control) succeeded while others (e.g. biological weapons ban, past climate pacts) faltered, focusing on the role of verification, transparency, and strategic coordination.
Ultimately, we will produce a comprehensive policy report (with tailored briefings for governments, the intelligence community, and AI governance researchers) that outlines practical steps to improve international cooperation in AI governance and reduce the risk of a catastrophic AI arms race.
Research tasks:
- Review historical analogs
- Identify AI-relevant CBMs
- Analyze which CBMs are needed for each provision of MIRI’s agreement
- Analyze bilateral and multilateral scenarios
- Compile findings into recommendations
Theory of change
Enables international cooperation on AI governance.
Your role
Mentees will execute the research tasks listed in the proposal above.
Prerequisites
Familiarity with international relations is preferred.
About the mentor
Robi Rahman
MIRI Technical Governance Team
Robi's work tracks the inputs and development of advanced AI systems. Before joining MIRI, he worked at Epoch AI, building the database of machine learning hardware, GPU clusters, and AI models, and investigating their costs, algorithms, and development process. Robi is a contributor to the Stanford AI Index, and has a master's degree in data science from Harvard University.