Hassabis, Altman, Amodei Unite on AI Regulation
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Hassabis, Altman, and Amodei each published detailed regulatory proposals within the past five weeks — the first time the Google DeepMind, OpenAI, and Anthropic CEOs are on record, in writing, converging on the same diagnosis and remarkably similar prescriptions for frontier AI.
- Hassabis's framework, published Tuesday on Substack, drew rare public praise across the bitterly competitive AI industry from rivals including Altman, Microsoft CEO Satya Nadella, and even longtime rival Elon Musk; Anthropic co-founder Jack Clark called it "excellent."
- All three CEOs agree on five framework points: independent third-party testing of frontier models before public release, a single governing body, U.S.-led (not state-by-state or foreign) authority, threat awareness around cyber and bioweapon capabilities, and targeting only the small class of frontier models capable of catastrophic or strategic risk — none is calling for a broad AI crackdown.
- The CEOs split on who should be the final referee: Amodei wants an "FAA for AI" with federal power to block a model's release immediately from Day 1; Hassabis wants a "FINRA for AI" — an industry-funded, federally overseen body starting with voluntary pre-release reviews; Altman pushes an "IAEA for AI" — a U.S.-led international forum that uses access to frontier models and markets as leverage for compliance.
- The Trump administration publicly champions AI deregulation and has resisted "an FDA for AI," but has already been forced into improvised regulation twice this summer — first restricting Anthropic's Fable and Mythos models, then delaying access to OpenAI's GPT-5.6.
- Critics warn the emerging framework could amount to regulatory capture: OpenAI, Google, and Anthropic already have the lawyers, security teams, government relationships, and technical staff to navigate a complex certification process, while startups and open-source developers would face a much steeper climb; Meta's Mark Zuckerberg is reportedly working on his own memo as well.
Why it matters: Two ad hoc Trump administration interventions this summer — over Anthropic's Fable and Mythos and OpenAI's GPT-5.6 — suggest the era of voluntary self-policing is already over, and now the three CEOs with the most to lose from an AI slowdown are lobbying hardest for the rules. Critics warn the framework could entrench OpenAI, Google, and Anthropic while locking out startups and open-source developers who lack the legal and technical staff to navigate certification.
