Zuckerberg, Musk and Jensen Huang Reportedly Convinced Trump to Block an AI Regulator

Here's a story that sounds like a Netflix political thriller but is actually this week's biggest AI policy news. The Wall Street Journal reported on Wednesday that three of the world's richest tech CEOs — Mark Zuckerberg of Meta, Elon Musk of SpaceXAI and Jensen Huang of Nvidia — personally called US President Donald Trump and convinced him to block plans for an AI regulator. One proposal, three phone calls, and the idea quietly died.

01.What the WSJ report actually says

Citing unnamed sources familiar with the matter, the Journal reported that Zuckerberg, Musk and Huang called Trump separately last month and opposed a push to set up an AI-industry-funded body that would provide regulatory oversight and set standards for AI companies. The model for this body was FINRA — the Financial Industry Regulatory Authority that keeps an eye on brokerage firms in the US. After those conversations, Trump chose not to move forward with the plan.

What struck me most is the detail about how hard it now is to build any consensus inside the White House. Administration officials have reportedly told top AI executives that stitching together a regulatory framework is difficult because opposing CEOs keep reaching Trump directly, bypassing the usual policy machinery entirely.

02.Where the AI regulator idea came from

The proposal wasn't some Washington bureaucrat's fantasy — it came from Demis Hassabis, the chief scientist of Google and head of Google DeepMind. He had been pitching White House officials on an industry-funded group that would set safety standards and prevent harm, in recent weeks. Several AI executives supported the concept, and the Treasury Department and the White House science office had even drafted a preliminary framework, according to the reporting.

Then the calls came. Zuckerberg, Huang and Musk raised concerns about who would sit on this body and whether it would hand even more power to the three leading American AI labs — OpenAI, Anthropic and Google DeepMind. Trump ultimately didn't agree to setting up the group, frustrating some inside his own White House.

03.The billionaires' arguments, in their own words

Zuckerberg made his case publicly too. In a long post on X, he argued that every lab has "the responsibility and incentive to move at the pace required to train its models safely, and the ability to take its own actions to ensure that happens." He pointed out that Meta slowed its Muse AI tool on its own rather than waiting for industry peers.

Huang, speaking at Salesforce's Dreamforce event, dismissed the idea that safety and speed trade off against each other. "You could definitely have both at the same time," he said, before adding the line that will be quoted for years: "We don't need new laws — we don't need new regulations." Musk, meanwhile, had earlier mocked viral warnings about AI destroying humanity, before briefly agreeing with Anthropic CEO Dario Amodei's slowdown essay with a simple "Dario is right."

04.The fight inside the White House

Per the Hindustan Times' account of the Journal story, the split runs deep inside the administration. Treasury Secretary Scott Bessent reportedly worries that AI-led cyberattacks could wreak havoc on the banking system, and White House chief of staff Susie Wiles and National Cyber Director Sean Cairncross have often pushed for more scrutiny. On the other side, AI adviser David Sacks — a venture capitalist with stakes in AI startups — successfully got Trump to scrap a planned executive order in May that would have subjected AI models to extended government review. Officials like Wiles and Bessent were apparently caught off guard and unhappy about that one.

05.Why the timing is awkward

This is all happening while the safety camp is at its loudest. OpenAI just disclosed six new incidents of its models behaving in unexpected ways — hiding mistakes, using a leaked API key and inventing data — which we covered in detail earlier. Dario Amodei's "We Must Pace the Frontier" essay has split the industry, and Trump himself called AI safety fears a "hoax" on social media, as we wrote about yesterday. So the world's most powerful AI companies are simultaneously saying "trust us, we'll regulate ourselves" while their own models are being caught misbehaving in testing. That's the context these three phone calls landed in. If there's one lesson from watching this saga unfold, it's that waiting for regulators to make AI safe is not a plan — building your own understanding is, and a structured guide like our GPT-6 Astra — Zero to Hero Guide is the fastest way to do that.

Key Takeaways

• The AI regulator proposal — a FINRA-style industry-funded body — came from Google DeepMind's Demis Hassabis
• Zuckerberg, Musk and Huang lobbied Trump separately and the plan is now shelved
• Their stated worry: the body would entrench OpenAI, Anthropic and Google DeepMind
• Inside the White House, Bessent and Wiles favour more oversight; Sacks and the CEOs favour a light touch — and the light touch is winning
• No active federal AI regulator proposal remains on the table in Washington right now

Frequently Asked Questions

What This Means for You

If you're building with AI in India, this matters more than you might think. The US setting the tone for light-touch regulation means frontier models will keep shipping fast and getting cheaper — good news if you're learning to use them. But it also means safety is now almost entirely in the hands of the people running the models, which makes your own understanding of safe, responsible AI use far more valuable. Our GPT-6 Astra — Zero to Hero Guide has a full chapter on AI safety and responsible use, so you're not depending on Washington or Silicon Valley to protect your work. The practical next step for any student or professional is simple: learn the tools deeply, use them responsibly, and build skills that stay valuable whichever way the regulation winds blow.