Behind the Curtain: AI godfathers converge on regulations

The CEOs of OpenAI, Anthropic, and Google DeepMind have converged on an unlikely consensus: frontier AI needs regulation, and soon. Per Axios, Sam Altman, Dario Amodei, and Demis Hassabis have each published detailed policy frameworks in recent weeks. They agree that independent third parties should test powerful models before release, that a federal body should oversee access, and that the U.S. should lead a globally coordinated approach. But they diverge sharply on implementation. Amodei wants an FDA-style agency with immediate power to block releases. Hassabis prefers an industry-funded standards body with federal oversight. Altman pushes an international forum modeled on the IAEA. Meanwhile, the Trump administration—publicly opposed to AI regulation—has already intervened twice this summer to restrict models from Anthropic and OpenAI. Critics worry that regulation, even with good intentions, risks entrenching the biggest players and locking out startups and open-source developers.

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton