Reddit
Anthropic, OpenAI and Google have been meeting since July about a private AI standards body, and disagree on what it is
CNN reported on 2026-09-14 that the three labs have held working-group meetings since July on an industry-led standards organization, and that their models of it differ sharply. Amodei wants an FAA for AI, a federal agency able to block a model release from day one; Hassabis wants a FINRA for AI, industry-funded and federally overseen, starting with voluntary pre-release reviews and staffed partly by open-source representatives; Altman reportedly thinks the labs will have to build it themselves absent government backing. The disagreement is the story, because it means the pacing call has no agreed enforcement mechanism behind it.
↳ Follow the thread