Three rival labs have been quietly writing their own safety rulebook.
The Information reported on Sunday that Anthropic, OpenAI and Google have met since July on a standards body. It would set shared tests every frontier model passes before release, plus outside evaluations. The three still disagree over how far government should be involved, and nothing is signed.
Why it mattersRight now the company that builds a model decides on its own when it is safe to ship. A shared rulebook could change that, or price smaller rivals out of the market.
Read in the app — free, in 9 languages
Related stories
Microsoft's new rulebook bars its models from hiding how they think.2026-09-15Nvidia's boss put Trump on speakerphone to reject an AI slowdown.
2026-09-15Beijing answered the AI slowdown plan by calling it fearmongering.
2026-09-14Cohere's boss calls the labs' safety plan a cartel in disguise.
2026-09-14Xi offered BRICS an open-source AI club that China would lead.
2026-09-14