Six governments and OpenAI, Anthropic sign AI safety declaration

The declaration did not emerge in a vacuum. Its text points to a chain of earlier efforts — the Hiroshima process, the Frontier Safety Commitments of the Seoul AI Summit, the G7 Evian AI meeting, the United Nations Security Council meeting on AI held on 23 September, and the "Call for Control of Frontier AI Models" — that the signatories say laid the groundwork for concrete cooperation.

The document's framing is notably collective. It argues that frontier models transcend national borders in development, deployment and impact, and that no single state, company or international organization can manage all the risks alone. That premise is what led the six governments and the named companies to sign the same page, alongside civil-society groups such as LawZero, SaferAI and Ifri.

What remains is operational. The signatories commit to establishing a structured framework for cooperation based on three strategic commitments, but the declaration does not yet spell out what those commitments require in practice, and it stays open to new endorsements as it is updated.