The White House has been working on a plan to create the first national AI regulator.
Before any of it went public, the president called one of the world’s most powerful tech executives to talk about it. The executive told him he was against it.
Meta CEO Mark Zuckerberg told President Donald Trump during the week of Aug. 17 that he opposed a proposal to establish a new federal AI oversight body, according to people familiar with the call.
Trump placed the call. The exchange, which has not previously been reported, shows how directly the biggest names in tech are shaping AI policy through private conversations at the top, Politico reported.
What the proposed AI regulator would actually do
The proposal has been championed by Nobel Prize-winning Google DeepMind co-founder Demis Hassabis, who wrote a July essay arguing the U.S. should create a new AI standards body modeled after the Financial Industry Regulatory Authority, or FINRA.
FINRA is the private, nonprofit body that writes and enforces rules for more than 3,000 brokerage firms and around 630,000 registered representatives. It is funded by member fees and operates under Securities and Exchange Commission supervision, with rule changes subject to SEC review.
More Mark Zuckerberg:
- Mark Zuckerberg, Microsoft CEO just made major AI decision
- Zuckerberg’s stock rebound made him world’s 5th richest at $222B
- Mark Zuckerberg backs Elon Musk Silicon Valley decision
The proposed AI version would work similarly. It would review advanced AI models, test them for potential risks and create common criteria for assessing systems before they are deployed more broadly.
Hassabis briefed White House officials about the concept this summer. White House officials had separately previewed the plan with Trump and with major AI companies, including Meta, OpenAI and Anthropic, in mid-August.
Supporters of the idea argue it would bring in the technical talent needed to conduct meaningful evaluations. It would also establish unified standards that currently do not exist.
Critics, mainly on the industry side, worry that voluntary pre-release review could eventually become mandatory. A model held up in testing is a model not generating revenue.
What Zuckerberg told Trump and why it matters
When Trump called Zuckerberg, the Meta CEO said he opposed the proposal. He did not ask Trump to reverse course. But he did say that anyone appointed to such a body should reflect Trump’s own preference for a light-touch approach to AI, according to a person familiar with the conversation.
A Meta spokesperson declined to comment. A White House spokesperson said the administration “is committed to balancing innovation and security in AI policymaking.”
The call is the second time this year that a tech executive has used direct access to Trump to shape federal AI policy. In May, former White House adviser David Sacks called Trump on the morning of a planned signing ceremony and persuaded him to cancel a sweeping AI executive order, according to Politico.
In August, Zuckerberg published an essay arguing that any policy slowing an AI model’s release, even by a month, would meaningfully damage U.S. competitiveness against China. His public position and his private call with Trump point in the same direction.
Why Sacks and Musk want a different model
Sacks, who served as Trump’s AI and crypto czar, has been openly critical of a government AI regulator. He has called the idea a “DMV for AI,” where models queue up waiting for approval.
His preferred alternative is a voluntary industry group modeled on the Motion Picture Association, which administers the film-rating system. Ratings like PG-13 and R are not laws. They are industry standards the studios adopted to head off federal content regulation.
“What the MPA did then was promote standards that then forestalled more intrusive, heavy-handed government action,” Sacks said on his All-In podcast in August.
Elon Musk has reportedly backed that approach as well. He did not respond to a request for comment.
On the other side, Anthropic co-founder Jack Clark posted favorably about the FINRA model on X in July. Anthropic declined to comment on its formal position. OpenAI and Google also did not respond to requests for comment.
What the debate means for Meta stock and AI investors
The regulatory question has real stakes for Meta specifically. The company has built its AI strategy around open-weight models, releasing the weights of its Llama systems publicly. A pre-release testing regime could complicate open-weight releases in ways that would not affect closed models from OpenAI or Anthropic the same way.
That commercial reality sits underneath Zuckerberg’s policy argument. He may be right that slower model releases hurt U.S. competitiveness against China. He also has a direct financial reason to prefer voluntary standards over a regulator with the authority to delay or block a release.
The proposal is still under active discussion in the White House. It has not been shelved because of Zuckerberg’s call. The White House has not indicated a timeline for a decision.
But the pattern is now clear. On AI regulation, the largest tech companies are not waiting for Washington to decide. They are calling the president directly to shape what gets decided.
That is a different policymaking process than the one described in civics textbooks. It is the one that appears to be operating.
Related: Mark Zuckerberg sends shocking message to Meta employees
