OpenAI and Anthropic employees are circulating a petition that calls on the US government to back an international mechanism to deliberately pace frontier AI advancement and prevent the technology from advancing too fast. The effort draws on statements from company leaders about the possible need to slow progress. Reports indicate more than one thousand signatories across labs including Meta and Google. Commenters note related concerns over extinction risks, adversarial actors, and the challenges of self-regulation by labs developing the systems.