Recent research
This workshop paper catalogs the compute used to develop AI algorithmic innovations used in open-source frontier models. We assess trends in the compute used in the experiments to develop innovations, and investigate the implications for compute caps.
Technical Requirements for Halting Dangerous AI Activities
This workshop paper discusses technical interventions that could contribute to a global halt on dangerous AI activities. We then provide a breakdown of which interventions play a role in various AI governance plans.
Response to RFC on AI Diffusion Rule
We respond to the Request for Comment about the AI Diffusion Rule. We discuss the interaction between export controls and AI risks, considerations for designing effective export controls, and specific changes to the Diffusion Framework.
Technical Governance Team Mission
AI systems are rapidly becoming more capable, and fundamental safety problems remain unsolved. Our goal is to increase the probability that humanity can safely navigate the transition to a world with smarter-than-human AI, focusing on technical research in service of governance goals.
governance goals:
Coordination
Strengthen international coordination to allow for effective international agreements and reduce dangerous race dynamics.
Security
Establish robust security standards and practices for frontier AI development, to reduce harms from model misalignment, misuse, and proliferation.
Development Safeguards
Ensure that dangerous AI development could be shut down, if there was broad international agreement on the need for this.
Regulation
Improve and develop domestic and international regulation of frontier AI, to prepare for coming risks, and identify when current safety plans are likely to fail.
Frequently asked questions