The Carolina Principles: Washington pushes back against new global AI rules at the G20
At a G20 technology meeting, a senior White House adviser urged member nations to avoid new AI regulations and new international oversight bodies, favouring research investment over binding rules — a stance that puts Washington on a different track from Brussels.
At a G20 technology gathering held in North Carolina, a senior White House technology adviser called on member nations to hold off on introducing new AI-specific regulations, presenting a set of principles — since nicknamed the “Carolina Principles” — that emphasise research investment and minimal government intervention over binding rules.
The US position explicitly opposes creating new international AI oversight bodies, arguing that fast-moving technology is poorly served by slow-moving multilateral institutions. That stance puts the United States on a noticeably different track from the European Union, which has spent the past two years building out the EU AI Act’s risk-tiered rulebook, and from countries pursuing their own domestic frameworks.
The disagreement matters beyond diplomacy. Where a country lands on the regulate-fast versus regulate-later spectrum increasingly shapes where AI labs choose to headquarter research efforts, where compute infrastructure gets built, and which markets get first access to the most capable new models — making G20-level positioning an early signal of where the next wave of AI investment may concentrate.