AI Regulation US urges G20 to avoid new rules for artificial intelligence

The United States has urged G20 countries to take a hands-off approach to artificial intelligence regulation and avoid introducing broad new rules, as technology leaders and commerce ministers gathered in North Carolina.

US tech adviser Michael Kratsios called on governments to adopt the “Carolina Principles”, which encourage countries to avoid creating entirely new AI regulations and instead develop targeted rules for novel situations involving the technology.

The approach largely reflects the interests of major US-based AI companies, which are seeking greater regulatory flexibility as they expand their businesses worldwide.

Kratsios, who is co-hosting the two-day meeting, said policymakers should not treat every emerging technology as a completely new policy challenge.

China has also signed the Carolina Principles, according to Kratsios. Chinese Technology and Science Minister Yin Hejun represented Beijing at the meeting, although the Chinese government had not publicly commented on the agreement at the time of the report.

The US position reflects growing concerns within Washington that strict regulations could slow the development and release of AI models, increase compliance costs and affect how companies design their systems to address safety and security concerns.

Tech leaders outline competing priorities

Several of the world’s leading technology executives addressed the gathering virtually.

Google DeepMind chief Demis Hassabis called for governments to establish safety testing for AI systems. Meta CEO Mark Zuckerberg, meanwhile, urged countries not to restrict open-weight AI models, whose core elements can be accessed publicly.

Dollar rises as Middle East tensions lift oil prices and revive rate-hike bets

Zuckerberg has increasingly positioned Meta as a major developer of open-weight AI systems, making regulatory restrictions a potential obstacle to the company’s strategy.

SpaceX CEO Elon Musk criticised European technology regulations, arguing that EU policies inhibit technological progress. He also called on countries outside China to develop additional energy sources to meet the growing electricity demands of AI data centres.

The expansion of data centres has become an increasingly important political issue in the United States, particularly ahead of the 2026 midterm elections, as communities debate their impact on electricity supplies, infrastructure and the environment.

Anthropic co-founder Tom Brown is also scheduled to address the G20 gathering.

The debate comes as concerns over AI safety continue to grow. A recent United Nations panel warned that technological advances are moving faster than scientific understanding and government policy.

Recent cybersecurity incidents involving AI agents have further intensified questions about the safeguards used to test and deploy increasingly autonomous systems.

US and China compete for AI leadership

The G20 meeting comes as the United States faces growing competition from China in the global AI race.

Chinese open-weight AI models have rapidly improved and are increasingly competing with proprietary systems developed by US companies such as Anthropic and OpenAI.

Their growing adoption by US companies has also raised national security concerns in Washington, particularly over the possibility of foreign interference in widely accessible AI systems.

Vivek Chilukuri, a technology and national security fellow at the Center for a New American Security, said the developments had increased pressure on the Trump administration to ensure that other countries remained within the US technology ecosystem rather than turning to alternatives.

The North Carolina meeting is part of a series of G20 gatherings being held in US cities ahead of the leaders’ summit in Miami in December.

OpenAI chief executive Sam Altman and Nvidia chief executive Jensen Huang are also expected to appear before delegates alongside US Commerce Secretary Howard Lutnick.

The meeting drew a small group of protesters outside the venue. Some criticised the expansion of data centres because of their energy requirements and potential environmental impact.

John Montavon, a protester from North Carolina, criticised Musk’s support for expanding data-centre infrastructure, arguing that its growing energy demand could contribute to higher global temperatures.