DooDooLamb News
OpenAI, Anthropic and Google secretly joined forces to collaborate on AI safety
Brief published September 17, 2026 ยท Original source published September 16, 2026
Original reporting by Mike Wheatley at siliconangle.com.
Automated brief. Verify important details at the original source.
What happened
OpenAI's global head of policy Chris Lehane stated that OpenAI has been collaborating with rivals Anthropic and Google's DeepMind to develop a shared framework for AI safety. The collaboration was not previously disclosed publicly, making this the first confirmed acknowledgment of coordinated safety work across three of the largest AI developers. The effort appears focused on devising common standards or guidelines, though the specific contents of any framework have not been released.
Why it matters
Cross-competitor coordination on safety standards could shape baseline practices across the industry. If the three organizations align on shared criteria, those criteria may effectively become de facto norms that influence how other developers, regulators, and deployers approach AI risk management. The secrecy around the collaboration raises questions about transparency and whether outputs will be made public.
What to watch
Whether a formal framework document is released, how other major AI developers respond to any proposed standards, and whether regulators treat this collaboration as a model for industry self-governance or as a reason to push for independent oversight.