US to urge hands-off AI regulation at G20, official says

The U.S. will press G20 members to take a hands-off approach to AI regulation and avoid creating new rules for the technology, according to prepared remarks seen by Reuters, at a gathering of industry titans and commerce ministers that begins on Tuesday in North ‌Carolina. U.S. tech adviser Michael Kratsios, who is co-hosting the two-day meeting, will advocate for countries to adopt the "Carolina Principles," according to a copy of his prepared remarks seen by Reuters. Countries that sign on agree to "reserve new regulation for novel considerations" when writing rules for AI. They also will invest in "foundational research to accelerate discovery" and strengthen commercial opportunities for new technology, according to the ‌prepared remarks.

Kratsios will tell the group that "policymakers do not need to approach each innovation in isolation and should not treat every emerging technology as a first-of-a-kind policy problem," according to ‌a copy of his prepared remarks seen by Reuters. Leaders are convening at a key moment in the AI sector's development and the global debate over how to govern it, and as the U.S. faces an increasing threat from China for AI supremacy.

Meta CEO Mark Zuckerberg will address the ministers by video chat, a U.S. Commerce Department spokesperson said, in lieu of company president Dina Powell McCormick. Google DeepMind co-founder Demis Hassabis was also added to the ⁠list of ​speakers who will address the ministers by video ⁠chat, a Google spokesperson said on Monday. Zuckerberg's speech Reuters previously reported that OpenAI CEO Sam Altman and Nvidia CEO Jensen Huang will separately appear before the delegates with Commerce Secretary Howard Lutnick, who is also hosting the meeting, on ⁠Wednesday. The U.S. plans to have SpaceX CEO Elon Musk and venture capitalist and former White House AI adviser David Sacks address the attendees live via video chat on Tuesday.

A United Nations panel recently warned that ​AI's developments are outpacing scientific understanding and government policy. Recent events, including a hack triggered by a rogue OpenAI agent that compromised the infrastructure of AI company Hugging Face, ⁠added to concerns about how closely leading AI companies are monitoring tests of their products. AI agents are programs that run with minimal human supervision.

The U.S. will press member countries not to set up new regulatory organizations to oversee AI development, ⁠a ​White House official said. The meeting is one of a series of gatherings taking place over the next few months in U.S. cities leading up to the G20 leaders' summit in Miami in December.

Chinese open-weight AI models, systems with publicly accessible core elements, are becoming increasingly capable when compared to proprietary systems from American AI giants such as Anthropic and OpenAI. Chinese-made open-weight ⁠models are gaining ground with U.S. companies, also posing a potential security risk if the country's leaders decide to interfere. "It has increased urgency for this administration to make sure that the ⁠rest of the world stays within the American ⁠tech ecosystem and doesn't seek alternatives," said Vivek Chilukuri, a technology and national security fellow at the Center for a New American Security.

Canadian officials are slated to attend, a spokesperson for the Canadian government said, adding that the delegation will advocate for countries to balance innovation with "public ‌trust and safety." The U.S. and ‌Canada are embroiled in a trade war, started by the U.S. president.