Paul Tudor Jones Warns US Is Dangerously Late on AI Regulation as Industry Support for Oversight Surges from 20% to 80% in One Year
Billionaire hedge fund manager Paul Tudor Jones has delivered a sharp public warning that the United States is dangerously behind on regulating artificial intelligence, telling CNBC that regulatory action is needed immediately: 'We should have already done it.' The intervention is notable because Jones is simultaneously bullish on AI as an investment — he told CNBC he recently bought more AI stocks — making his call for oversight a market participant's warning rather than a purely political one. His most striking data point: at a recent conference of AI experts and model makers, 80% of participants supported AI regulation, up sharply from approximately 20% a year earlier. One major AI company CEO reportedly said they were surprised the industry was not yet regulated. Jones specifically highlighted the need for watermarking — a technical mechanism to tag AI-generated content so it can be distinguished from human-created material — as a priority, particularly to combat deepfakes. The policy backdrop includes the EU AI Act (passed in 2024), the White House's national AI policy framework (published March 2026), and ongoing US-China discussions about AI safety guardrails ahead of a potential Trump-Xi summit. Jones also noted that the US and China are in discussions about establishing official AI safety dialogue, arguing the two countries should collaborate rather than allow their AI rivalry to become a crisis.
Why this matters
The rapid shift in industry sentiment — from 20% to 80% supporting regulation in a single year — is a leading indicator that self-regulatory and government intervention frameworks are moving from aspiration to near-term reality. For legal practitioners, the EU AI Act is already law and its high-risk provisions are now scheduled for December 2027 (delayed from the original timeline); the White House framework signals US federal legislation is in prospect. When regulation crystallises, it generates immediate compliance advisory demand: AI governance policies, data processing agreements, vendor due diligence on AI tools, and regulatory impact assessments for high-risk AI deployments. Watermarking requirements, if enacted, would directly affect financial services firms' use of AI-generated research and communications — a live issue for the City clients of Magic Circle and US firms.
On the Ground
A trainee supporting an AI regulatory advisory matter would draft AI governance policy documents aligned to the EU AI Act's risk classification framework, mark up data processing agreements for AI vendor relationships, and prepare regulatory impact assessment memos mapping client AI deployments against emerging compliance obligations. Vendor due diligence questionnaires for AI tool procurement would also fall within the trainee's scope.
Interview prep
Soundbite
When 80% of AI builders want regulation, the compliance advisory wave for City firms is no longer theoretical — it is imminent.
Question you might get
“How does the EU AI Act's risk classification system work, and what obligations would it impose on a UK financial services firm deploying an AI tool for credit underwriting decisions?”
Full answer
Paul Tudor Jones has warned publicly that the US is dangerously late on AI regulation, citing a surge in industry support for oversight from 20% to 80% of participants at a recent AI conference. The commercial significance is that this rapid shift in industry sentiment, combined with the EU AI Act already on the statute book and a White House framework published in March 2026, means AI-specific legal compliance work is transitioning from a niche to a mainstream practice area. For City firms advising financial services clients, the immediate implications include reviewing AI governance frameworks, updating data processing agreements with AI vendors, and preparing for potential watermarking or disclosure obligations. The US-China dimension Jones raised — potential official AI safety dialogue at a Trump-Xi summit — adds a geopolitical layer that will affect how multinationals structure their AI governance across jurisdictions.
My notes
saved