The United States has secured G20 consensus for the Carolina Principles for Emerging Technologies, placing an innovation-focused approach at the centre of the group’s policy discussion on artificial intelligence and other fast-moving technologies.
At a two-day Innovation Ministerial in Chapel Hill, North Carolina, on September 2nd, the framework was agreed, and the principles ask governments to use existing sector rules where they fit and reserve new regulation for situations that current laws cannot adequately address.
The agreement gives the U.S. a shared G20 statement supporting a lighter, more targeted route to technology policy, while allowing every member to build its own domestic framework.
What The Principles Set Out
While the document covers emerging technologies broadly, AI is still a major part of the ministerial agenda. It calls for governments to fund foundational research, work with private companies and universities, and create clearer paths from research laboratories to commercial products.
It also backs supervised regulatory sandboxes, which let organisations test technology under defined conditions. The principles also call for faster permitting and approval processes for pilot projects, real-world testing facilities, and public procurement that can help create early markets for new technology.
The agreement says governments should support secure, accessible, trustworthy, and interoperable infrastructure. It also calls for digital-literacy programmes and technical training developed with private-sector and other relevant partners.
A Targeted Approach to AI Rules
The most closely watched section deals with how countries should govern new technology after it reaches the market. The Carolina Principles advise governments to apply existing sector-specific frameworks where appropriate, and this could mean using rules and regulators already responsible for areas such as health, finance, transport, or consumer protection when AI is used in those fields.
Where existing frameworks leave a genuine gap, the text supports new regulation focused on that specific issue. It says new rules should address gaps instead of repeating protections that already exist.
The document does not set one legal regime for all G20 members, as it says the principles should be applied flexibly, recognising different national circumstances and legal systems. Members also reaffirmed their right to develop their own policies and preserve national sovereignty in governing emerging technologies.
A Surprising Show of Support
The agreement also brought an unexpected point of agreement between the U.S. and China. Chinese officials backed the principles despite the two countries’ wider competition over AI, with President Donald Trump and Chinese President Xi Jinping also expected to discuss technology issues in upcoming talks.
The ministerial brought major AI figures into the conversation. NVIDIA CEO Jensen Huang and OpenAI CEO Sam Altman attended in person, while Elon Musk joined by video. Google DeepMind CEO Demis Hassabis was among the notable voices calling for stronger AI safety measures.
The discussion came shortly after the European Union moved ahead with further enforcement of its AI Act in early August, highlighting the different approaches major economies are taking to AI regulation.
The Broader G20 Package
The Carolina Principles were one of six pillars in the G20 Innovation Ministerial Statement. The other pillars cover technology and prosperity, technical workforce development, AI intellectual-property policies, AI standards, and industrial innovation and supply chains.
The meeting was hosted by the U.S. Department of Commerce and the White House Office of Science and Technology Policy, with the Commerce Secretary Howard Lutnick saying the agreement promotes domestic manufacturing and intellectual-property frameworks that foster and protect innovation. White House science adviser Michael Kratsios also said the agreement reflects an approach built around flexible policies designed to promote innovation.
For the tech industry, the immediate outcome is a shared policy reference point across the G20. The published text leaves implementation to individual members through their own rules, funding programme and regulatory decisions.
