Zuckerberg Privately Called Trump to Oppose AI Regulator

Zuckerberg reportedly spoke privately with Trump in August, expressing opposition to a proposal to create a mandatory AI regulatory body modeled on FINRA. The White House is currently divided over the regulatory path, torn between a government review model similar to FINRA and a voluntary rating model based on the Motion Picture Association (MPA), with the final policy direction still unclear.

The direction of U.S. AI regulatory policy is being profoundly shaped by a series of private interactions between tech giants and White House officials.

According to Business Insider on the 4th, citing informed senior White House officials, Meta CEO Zuckerberg spoke with Trump during the week of Aug. 17, clearly expressing opposition to the proposal to establish a national AI regulatory body. The White House later responded: "The Trump administration is committed to balancing innovation and safety in AI policy development." The disclosure of this call once again reveals the reality of tech industry executives bypassing conventional policy channels to directly influence the president.

The regulatory body proposal is still under discussion within the White House and has not been shelved due to Zuckerberg's opposition. But this episode shows that policy directions already formed within the White House are constantly facing private lobbying pressure from industry—pressure sufficient to slow or even redirect advancing plans.

Proposal Origin: Modeled on FINRA

The core proponent of this regulatory concept is Nobel laureate and Google DeepMind scientist Demis Hassabis. In an article published in July, he suggested that the U.S. should establish a new AI "standards body" modeled on the financial industry regulator FINRA, specifically to address AI risks such as cybersecurity threats. According to a government official familiar with the matter, Hassabis briefed White House officials on the concept this summer. Google did not respond to a request for comment.

FINRA is a private, non-profit membership organization operating under the oversight of the U.S. Securities and Exchange Commission (SEC), responsible for writing and enforcing rules for more than 3,000 broker-dealers and 630,000 registered representatives. Its operating funds come from member fees, and rule changes must be reviewed by the SEC.

According to White House officials, senior White House officials previewed the FINRA-style regulator plan to Trump and major tech companies including Meta, OpenAI, and Anthropic in mid-August. Under the White House's vision, the body would review and risk-test AI models before deployment. Supporters argue this model could both bring in sufficient technical talent to conduct testing and establish unified standards for evaluating AI models.

Zuckerberg's Opposition: Regulatory Direction Should Align with Trump's Philosophy

Zuckerberg did not publicly voice opposition but chose to express his position through private channels. The report, citing another informed source, said Zuckerberg did not ask Trump to change his stance but told him that if personnel were appointed to a regulatory body in the future, they should be people aligned with Trump's own AI philosophy—which is widely seen as favoring light regulation. The source also said the call was initiated by Trump.

In a public article in August, Zuckerberg had already expressed doubts about strict government regulation. He argued that government and AI companies should work closely together, but any policy that could delay model releases—even by a month—would "pose a significant risk to America's leadership relative to China."

A Meta spokesperson declined to comment.

Internal Divisions: Two Paths Undecided

The call reflects deep divisions within the White House over the AI regulatory path. Officials are currently weighing two directions: one is to establish an industry regulator modeled on FINRA; the other is to create a voluntary industry organization modeled on the Motion Picture Association (MPA)—a plan pushed by David Sacks, Trump's former AI and crypto advisor.

Sacks is strongly opposed to a government regulator, likening it to an "AI DMV" where models would "queue up for review." According to Politico, in May, Sacks persuaded Trump with a single phone call to cancel the signing of the first AI executive order on the day of the ceremony.

On the "All-In" podcast aired on Aug. 21, Sacks outlined his proposal, developed with support from Musk: an AI rating system modeled on the MPA, providing voluntary labels for AI models similar to movie ratings. "What the MPA did was promote standards, thereby preventing more invasive, heavy-handed government intervention," he said. Musk did not respond to a request for comment.

Industry Cautious, Policy Direction Unclear

Among major AI companies, none has publicly expressed support for or opposition to the regulatory proposals. Anthropic co-founder Jack Clark posted on X in July acknowledging the concept, but Anthropic declined to comment when asked about its official position; OpenAI and Google did not respond to interview requests.

Critics have concerns about the FINRA-style model, arguing that voluntary standards and pre-deployment testing requirements, once established as practice, could eventually be codified into mandatory regulations.

This policy debate is occurring at a critical juncture of rapid AI capability advancement—AI agents can already autonomously browse the web, manipulate software, and write code with limited human oversight, and have exhibited loss of control in several high-profile incidents. In the hands of adversaries, these systems could also identify vulnerabilities or launch cyberattacks at speeds exceeding existing defense systems.

Trump's one-on-one meetings with tech executives have become an informal but significant variable in the White House's AI policy-making process—one that can at any time slow or redirect carefully advanced internal plans, leaving some senior officials feeling constrained.