Washington’s proposed “Carolina Principles” turn AI regulation into a competitiveness strategy: fewer new rules, greater reliance on existing regulators, and more room for American technology companies to expand.

The United States is asking the world’s largest economies to avoid building broad new regulations or regulatory bodies specifically for artificial intelligence. At a G20 technology gathering in Chapel Hill, North Carolina, White House science and technology adviser Michael Kratsios is promoting what the administration calls the Carolina Principles. The framework would encourage governments to reserve new regulation for genuinely new problems that existing law cannot address, increase investment in foundational research, and rely more heavily on sector-specific oversight and cooperation between governments and technology companies. The proposal is framed as regulatory philosophy. It functions as economic competition.

Rules Are Becoming Industrial Policy

The companies with the greatest exposure to global AI regulation are overwhelmingly American. OpenAI, Nvidia, Meta, Google and other U.S. technology companies are spending extraordinary amounts of capital building models, chips and computing infrastructure, and Nvidia chief executive Jensen Huang and OpenAI chief executive Sam Altman are among the leaders attending the G20 gathering. Rules imposed abroad reach American companies whether or not Washington enacts anything comparable at home.

That changes what regulation is worth strategically. A country requiring lengthy approvals, extensive testing or new compliance structures may slow the deployment of AI systems inside its own economy. A country that lets companies move faster may attract investment and accelerate adoption while accepting greater safety, competition and consumer risk in exchange. Washington has made clear which side of that tradeoff it favors.

The Carolina Principles would discourage dedicated AI regulators and push governments toward existing institutions unless the technology creates genuinely novel problems. Supporters argue that healthcare, financial services, transportation and other regulated industries already have agencies capable of addressing harms inside their sectors, and that a new layer of AI bureaucracy would duplicate oversight and slow innovation.

China Changes the Calculation

The policy arrives as Chinese open-weight AI models have become more competitive with proprietary American systems. Reuters reports that U.S. officials increasingly treat those advances as a national-security and economic-competitiveness concern rather than a technology story. Regulatory policy has become part of a broader contest over whose technology ecosystem becomes the default.

If businesses, governments and developers across emerging markets build on American chips, cloud infrastructure and models, U.S. companies gain commercial reach and Washington retains influence over the underlying architecture. If Chinese models become cheaper, easier to deploy or less constrained by American export rules, that influence gets harder to hold. A lighter global regulatory environment can serve American industry without ever naming an American company.

Not every G20 government shares that emphasis. Canada told Reuters that innovation has to be balanced against public trust and safety, reflecting the competing concern that increasingly capable systems may create risks existing regulations were never designed to handle. The Carolina Principles are also not G20 policy. They are a U.S. proposal advanced during the American G20 presidency, not an agreement member governments have accepted.

That distinction matters, and so does the direction of the argument. The global AI contest is no longer a question of which country has the best model or the most advanced chip. Countries are now competing over the environment in which those technologies will be allowed to operate, and Washington is betting that fewer new rules will let American companies move faster. AI deregulation has become more than domestic policy. It is an export strategy.