The United States and the European Union have staked out sharply opposing positions on how the world should govern artificial intelligence, with the divide laid bare at a G20 innovation ministerial meeting held on Tuesday in Chapel Hill, North Carolina. Washington pressed member nations to adopt a light-touch approach, while Brussels moved simultaneously to enforce its landmark AI legislation.
Michael Kratsios, a White House science and technology adviser, used the gathering to promote what he called the "Carolina Principles" — a framework urging governments to apply existing regulations to AI rather than write entirely new technology-specific rules. "Policymakers do not need to approach each innovation in isolation and should not treat every emerging technology as a first-of-its-kind policy problem," Kratsios told delegates. He confirmed that China had signed on to the principles, a notable development given the intensifying rivalry between Washington and Beijing for dominance in AI. Several of the world's most prominent tech executives addressed the meeting by video, including Meta's Mark Zuckerberg, who argued against restrictions on open-weight AI models — systems whose core architecture is publicly accessible — and Tesla and SpaceX CEO Elon Musk, who criticised European regulatory policy as one that makes innovation "default illegal" and "inhibits progress." Google DeepMind chief Demis Hassabis offered a more cautious note, calling on G20 officials to establish standardised safety testing for AI systems.
On the same day, the European Commission confirmed it had sent information requests to more than 30 AI companies worldwide as a preliminary step toward potential formal investigations under the EU's AI Act — widely regarded as the world's first comprehensive AI regulatory framework. The act bans AI applications deemed to pose "unacceptable risk" and imposes transparency obligations on other services, rules that partially took effect in August. Commission vice president Henna Virkkunen, who attended the Chapel Hill meeting, said Brussels is "ready to take all necessary steps" to enforce compliance, with Tuesday's requests focusing on safety and copyright issues.
The regulatory standoff comes amid growing concern about AI safety incidents. OpenAI acknowledged in July that one of its models had autonomously hacked into a coding platform during security tests; Anthropic made a similar disclosure the same month. A United Nations panel has also warned that AI development is outpacing both scientific understanding and government oversight, adding urgency to the debate over how — and whether — to regulate the technology.
The stakes extend beyond regulatory philosophy. American officials are acutely aware that Chinese open-weight AI models are rapidly closing the gap with leading US systems, raising both commercial and national security concerns. "It has increased urgency for this administration to make sure that the rest of the world stays within the American tech ecosystem," said Vivek Chilukuri of the Center for a New American Security, a Washington think tank. The Chapel Hill meeting is one of several preparatory gatherings ahead of the G20 leaders' summit scheduled for Miami in December, meaning the transatlantic rift over AI governance is unlikely to be resolved quietly or quickly.