Global AI rules are fragmenting
The White House's AI blueprint is already facing pushback in Congress and looks set to clash with state authorities, while Asia and India are rolling their own high‑impact rules—India even launched 'AI Sutras' with a redressal mechanism. The result is regulatory patchwork firms must navigate now, not later. (wwmt.com) (aicerts.ai)
The White House published its National Policy Framework for Artificial Intelligence on March 20, 2026, a legislative recommendations document that sets six policy objectives and asks Congress to streamline data‑center permitting and limit “open‑ended” liability for developers. (whitehouse.gov) House Democrats and some progressives immediately criticized the framework as insufficiently protective, with Rep. Valerie Foushee calling it “falling short” of people‑first safeguards on March 20, 2026. (foushee.house.gov) Senate Republicans are simultaneously advancing federal preemption as a feature of national AI bills — Senator Marsha Blackburn released a draft framework this month aimed expressly at overriding at least some state AI laws, intensifying intra‑GOP debate over a uniform federal standard. (news.bgov.com) State activity remains extensive: a Transparency Coalition report counted 73 new state AI laws enacted in 27 states during 2025, with California (13 laws) and Texas (8 laws) among the largest movers, creating multiple operational regimes companies must track. (transparencycoalition.ai) Across Asia, jurisdictions are diverging: South Korea’s AI Basic Act and Japan’s framework set statutory and institutional obligations for “high‑impact” systems, while China has layered algorithm‑recommendation, AIGC labelling and deep‑synthesis rules plus national standards that took effect in late 2025. (gdprlocal.com) India’s India AI Governance Guidelines (the seven “Sutras”) formalize a principle‑based model and create new bodies — the AI Governance Group, Technology & Policy Expert Committee and an AI Safety Institute — while calling for grievance/redress mechanisms and transparency reporting for harms. (static.pib.gov.in) Regulatory divergence is already carrying enforcement‑level detail: China’s measures require algorithm filings and AIGC labelling, India’s guidelines mandate institutional oversight and redress steps, and the U.S. proposal seeks federal preemption — a set of mismatched obligations that vendors and deployers must map across jurisdictions. (whitecase.com)
