AI regulation fight erupts as Trump calls AI risk a hoax

Why the White House is betting on frontier labs to police themselves — and what it means for businesses building with AI

Washington’s fight over AI regulation broke into the open this week, and it did not stay polite. On Bloomberg Television’s Balance of Power, President Trump dismissed fears that artificial intelligence could take over the world, calling the idea a hoax in a comment blunt enough to become the episode’s headline. Minutes later, his own AI policy chief appeared on the same program making a far more careful case for why Washington should still hold back on new rules.

Trump Calls AI Existential Risk a Hoax

The episode opened with Bloomberg’s Jeff Mason reporting on the President’s pushback against AI safety concerns — the same warnings that executives at Anthropic and OpenAI have been raising through much of 2026 about frontier models advancing faster than the guardrails around them. As the episode’s own title put it, Trump considers predictions of AI taking over the world to be a hoax. For an administration that has otherwise leaned into AI as an economic and geopolitical priority, the message to Silicon Valley was clear. Slow down on the doom talk, not the deployment.

David Sacks Makes the Case for Letting Labs Police Themselves

The more substantive argument came from David Sacks, the White House’s point person on AI policy and co-chair of the President’s Council of Advisors on Science & Technology. Sacks told Bloomberg that if Anthropic and OpenAI genuinely believe their most advanced models carry serious risks, the responsible move is for those companies to slow their own development — not to wait for Washington to legislate. It puts the burden of caution on the labs themselves rather than on a regulator.

No New Rules, But Slow Down If You Believe the Risk

Sacks was explicit that he does not think new AI-specific regulation is necessary for labs to act on their own risk assessments. In his framing, if a company’s internal safety review concludes a model is too dangerous to ship, that is a corporate governance decision, not a legislative one. It is a notably different posture from the audits-and-oversight model many AI safety advocates favor, and it keeps the decision inside the labs rather than inside a regulatory agency.

The Regulatory Capture Warning

Sacks also raised a concern that has circulated in tech policy circles for more than a year: that heavy-handed AI rules risk becoming a form of regulatory capture, where the largest, best-resourced labs help shape the rules that then lock out smaller competitors. Pressed directly on AI transparency, audits and oversight, Sacks’ answer amounted to a bet that competitive pressure and reputational risk will keep frontier labs disciplined more effectively than a new compliance regime would.

The US-China AI Race Is the Real Constraint

Underneath the domestic argument sits a geopolitical one. Sacks tied the regulation question directly to U.S.-China competition on AI, the lens through which this administration evaluates almost any AI policy proposal. If Washington slows its own labs with new rules while Beijing does not, the argument goes, the U.S. risks ceding ground in a race it treats as strategically decisive. That calculus is why founders and technology leaders should expect the current light-touch approach to hold for now, whatever individual lawmakers say heading into the midterms — the same Bloomberg panel later noted that AI regulation is already becoming a live issue on the campaign trail.

None of this is happening in a vacuum for businesses already building with AI. The pace of frontier model releases that Sacks is defending is the same pace that has made tools like AI coding agents and systems built on a million-token context window viable for small teams almost overnight. A slower, more heavily audited industry would likely mean slower access to those capabilities too.

What Founders and Business Leaders Should Take From the Debate

For entrepreneurs and small business leaders watching from outside Washington, the practical takeaway is not which side of the hoax argument is right. It’s that no binding AI regulation is coming from this administration in the near term, and that has three concrete implications. First, the compliance burden for adopting AI tools in your business will keep resting on your own judgment and vendor selection rather than a government checklist, so due diligence on any AI vendor’s safety and data practices matters more, not less. Second, the pace of capability improvement Sacks is defending will likely continue, so the competitive advantage of adopting AI early is unlikely to shrink soon. Third, the regulatory capture warning cuts both ways for smaller companies — it is worth watching whether any future rules end up favoring incumbents with existing compliance teams over leaner competitors.

The clearest lesson from Sacks’ appearance is one familiar to anyone who has pitched a business case internally: proof beats pitching. Sacks is effectively betting that labs demonstrating restraint will earn more trust than any government mandate could impose — and that bet is now the working assumption behind AI regulation in the United States for at least the rest of this term.

Slow down on the doom talk, not the deployment.

Sacks is effectively betting that labs demonstrating restraint will earn more trust than any government mandate could impose.

▶

Sources

• Balance of Power — Trump: AI 'Taking Over the World…is a Hoax' (Bloomberg Television) — David Sacks' comments on AI regulation, frontier labs, regulatory capture and the US-China AI race appear starting at 21:50 in the episode.