End-User Safety Regulat!on for Frontier AI

This policy establishes a practical end-user safety regime for frontier AI models: highly capable general-purpose AI models or systems that can perform a wide variety of tasks and match or exceed the capabilities of the most advanced models. It converts global norms into enforceable duties for developers, deployers, and intermediaries, prioritising safe, secure, trustworthy, human-centric, and rights-respecting AI in line with the OECD AI Principles, the Seoul Declaration, and the Hiroshima AI Process.