Third-party audit standards
Developing third-party audit standards for high-stakes AI deployments in public services — verifiable claims about model behaviour before benefits and lives depend on them.
Public servicesHigh-stakes AIAudit
Governance frameworks for autonomous systems and algorithmic constitutionalism.
αDeveloping third-party audit standards for high-stakes AI deployments in public services — verifiable claims about model behaviour before benefits and lives depend on them.
Evaluating the benefits and risks of open model weights for competition, security, and innovation policy.
Studying hardware-based controls as a lever for managing frontier model development and proliferation.
Proposing legal frameworks that allocate responsibility across the AI supply chain — from developers to deployers.