Twelve governments and ten AI companies including OpenAI, Anthropic and Microsoft sign the Paris Call for AI Safety and Security

Élysée · Declaration

Read the source (opens in a new tab)
What happened
The Élysée published the Paris Call for AI Safety and Security on November 9, 2026, signed that day by 12 governments, 10 AI companies and 7 civil society groups.
Key points
  • Signers commit to a common framework for evaluating frontier-model risks, focused on cyber and CBRN misuse and on "growing autonomy or model misalignment, including loss of control."
  • They agree to publish non-sensitive evaluation findings in a standardized format and to create a publicly accessible shared repository.
  • Developers commit to disclose capabilities, limits and known risks, support qualified independent evaluators where appropriate, and implement "fail-safe mechanisms to shut down model operations within their control."
  • Signers recognize the value of strengthening reporting arrangements to share timely information on "serious cyber and AI control incidents," and governments commit to confidential channels for whistleblowers raising systemic risks.
Evidence
A signed political declaration. Its commitments are stated as intentions, some qualified "where appropriate," with no enforcement mechanism or deadlines. Company signers include Cohere, Hugging Face, Mistral AI and Sakana AI.
Uncertain
The page says the statement "remains open to new endorsements and is being updated," so the list of signers may change.

See it in the latest record