Quick answer: OpenAI’s Chris Lehane confirmed on September 15, 2026 that OpenAI, Anthropic, and Google DeepMind have spent several weeks discussing the creation of an industry-wide AI safety standards body, including plans to embed independent third-party evaluators inside frontier labs to monitor advanced models.
What happened
Speaking with U.S. lawmakers, Lehane confirmed the coordination follows Anthropic CEO Dario Amodei’s essay calling for the industry to slow the pace of frontier AI development and avoid catastrophic risks. Leaders including Sam Altman and Demis Hassabis have publicly endorsed the collaborative approach, and OpenAI has backed provisions in the FRONTIER Act that would require independent verification organizations inside frontier labs.
The effort faces political friction: the Trump administration has dismissed safety-driven slowdowns, with advisors warning that added regulation could disadvantage the U.S. against China in the AI race.
Why it matters
This is one of the most concrete moves yet toward formal, cross-lab AI safety governance rather than each company setting its own internal rules. It comes in the same week a class-action lawsuit accused these same companies of illegally coordinating to slow AI development — meaning the labs’ own safety collaboration is now being used as evidence in litigation against them.
Key facts
- Confirmed September 15, 2026 by OpenAI’s Chris Lehane
- Participants: OpenAI, Anthropic, Google DeepMind
- Proposal includes independent evaluators embedded in frontier labs
- OpenAI backs FRONTIER Act provisions requiring independent verification organizations
FAQ
Is this a government-mandated body?
Not yet — it is currently an industry-led discussion among OpenAI, Anthropic, and Google DeepMind, though it intersects with proposed legislation like the FRONTIER Act.
Why is this controversial?
Critics argue that safety-motivated coordination between competitors can also function as anticompetitive behavior, a tension now playing out directly in a related antitrust lawsuit.
Related reading
- Antitrust Suit Accuses OpenAI, Anthropic, SpaceXAI and Google of an Illegal AI Slowdown Pact
- OpenAI Shelves Its 2026 IPO, Citing Safety Concerns
- More AI industry coverage
Source: TechCrunch
