Bletchley Declaration on AI Safety — November 1, 2023
28 countries and the EU endorsed the Bletchley Declaration to cooperate on managing frontier AI risks, share safety research, and support international standards.
The declaration highlights the need to understand frontier AI capabilities, evaluate systemic risks, and align on transparency and accountability measures. Signatories agreed to convene future safety summits and build interoperable governance frameworks.
- Risk assessment. Countries will develop shared scientific understanding of frontier AI risks, including biological, cyber, and societal impacts.
- International collaboration. The declaration encourages information sharing between AI safety institutes and regulators.
- Standards development. Signatories support international standards bodies working on assurance and testing.
Global governance teams should track follow-up summits and technical working groups spawned by the declaration.