Global AI rules split as EU and U.S. diverge

AI regulation is moving away from a single global model, with governments pursuing different mixes of binding laws, executive action, agency guidance, sector rules and voluntary frameworks. The EU has built a comprehensive risk-based regime, while the U.S. lacks a comparable federal statute and continues to rely on executive policy, existing enforcement powers and an expanding set of state laws.

Several EU AI Act obligations are already live. Prohibited-practice rules have applied since February 2, 2025, general-purpose AI model obligations since August 2, 2025, and AI Office enforcement powers plus Article 50 transparency duties took effect on August 2, 2026. High-risk system obligations were delayed under the Digital Omnibus, giving standalone Annex III systems until December 2, 2027 and embedded Annex I systems until August 2, 2028.

In the U.S., Executive Order 14365 directs agencies to challenge state AI laws and seek a uniform federal framework, but no preemptive federal AI law had passed as of September 2026. At the G20 Innovation Ministerial on September 1-2, 2026, the U.S. promoted the non-binding Carolina Principles, favoring lighter regulation. Enterprises now need system-level inventories, jurisdiction mapping, vendor oversight and documentation to manage diverging rules.