Zuckerberg, Musk, and Huang reportedly talked Trump out of an AI FINRA

Three of the most powerful CEOs in technology reportedly dialed the White House and helped kill a plan that would have given artificial intelligence its own industry-funded watchdog. According to the Wall Street Journal, as summarized by Forbes, Meta's Mark Zuckerberg, SpaceX and xAI's Elon Musk, and Nvidia's Jensen Huang each spoke with President Donald Trump last month and convinced him to block efforts to create a FINRA-style standards body for frontier AI. The proposal, popularized by Google DeepMind's Demis Hassabis, would have sat somewhere between self-regulation and government oversight. After those calls, the Journal reported, the president did not move the plan forward.

A FINRA for models, not brokers

Hassabis has argued publicly for a U.S.-led standards body modeled on the Financial Industry Regulatory Authority: industry-funded, government-overseen, staffed with independent technical experts, and charged with testing advanced models before wide deployment. CNN and earlier reporting said OpenAI, Google, and Anthropic had discussed versions of such a body. Supporters frame it as a way to check cyber, biological, and deception risks without waiting for slow statutes. Critics hear something else: a club that certifies the labs that already dominate the frontier.

That is the fear Forbes attributes to Zuckerberg, Musk, and Huang. They worried the structure would entrench OpenAI, Anthropic, and Google DeepMind, the three American labs most associated with the biggest models. Administration officials, the Journal said, have told AI executives that White House consensus is hard because CEOs who dislike a proposal simply call Trump themselves.

Safety rhetoric meets market power

The episode sits on top of a wider split. Zuckerberg and Huang have recently opposed calls to slow frontier development. Musk has at times backed pacing arguments associated with Anthropic's Dario Amodei, yet still joined the push against this particular regulator. The odd result is a temporary alliance of rivals who disagree on "go fast" versus "go careful," but agree that an industry-funded standards board designed by the leading labs looks like regulatory capture dressed as safety.

For everyone outside the Mag 7 orbit, the story is less about three phone calls and more about who gets to write the rules. If the only people who can reach the president are the CEOs of Meta, Nvidia, and Musk's constellation, then AI governance becomes a private negotiation among platform owners. If Hassabis-style institutions only advance when OpenAI, Anthropic, and DeepMind like the shape of the table, then "sovereign" oversight risks becoming a franchise of the same three brands. Either path leaves smaller labs, open-source projects, and foreign competitors arguing over a map they did not draw. The Journal's account suggests that, for now, the map stays blank because three executives preferred it that way.