The United States is pressing G20 countries toward a lighter approach to AI regulation, arguing that governments should avoid broad, technology-specific rules that could slow development.

The proposal was discussed at a G20 technology meeting in Chapel Hill, North Carolina. U.S. officials promoted a framework known as the Carolina Principles.

The framework is a U.S. policy position. It is not a finished global rule, and it does not bind other governments to rewrite their laws.

What the Carolina Principles favor

The principles generally support rules that focus on particular risks and situations, rather than a new, sweeping set of laws written only for artificial intelligence.

That stance fits a wider U.S. effort to encourage rapid AI development and investment. American technology companies have also warned that heavy regulation could make it harder to build and deploy new AI products.

Several technology leaders took part in discussions on AI policy and infrastructure. The meeting included figures from companies such as Meta, Google DeepMind, and SpaceX.

Governments do not share one rulebook

Countries do not have a single position on how AI should be governed.

The European Union is taking a more detailed regulatory path. The EU’s AI Act introduces requirements covering transparency, safety, and the use of high-risk AI systems.

The contrast is becoming clearer. Washington is stressing innovation and fewer new barriers. European regulators are continuing to write and enforce rules intended to limit AI-related risks.

What the G20 talks can and cannot decide

The debate is growing as AI systems become more capable and spread through software development, business, education, healthcare, finance, and other industries.

Talks at the G20 can influence how governments think about AI policy in the years ahead. They do not create a single international statute. Each government will still decide how artificial intelligence is regulated at home.

For now, the U.S. proposal should be read as an argument about how to balance speed and safety — not as a final global settlement.