Trump Administration Unveils National AI Framework to Preempt State Laws

Donald Trump released the National Policy Framework for Artificial Intelligence on March 20, 2026, establishing a pro-innovation blueprint for congressional legislation. The framework advocates for a light-touch, sector-specific regulatory approach using existing federal agencies and regulatory sandboxes rather than creating a new centralized AI regulator. A primary goal of the plan is the federal preemption of state-level AI laws, which the administration argues create a fragmented patchwork that stifles innovation and threatens U.S. global competitiveness.

The framework is built around six to seven guiding pillars: protecting children and empowering parents, safeguarding communities against AI-enabled scams, respecting intellectual property, preventing government censorship, ensuring American AI dominance, and developing an AI-ready workforce. Notably, the administration asserts that training AI models on copyrighted material does not violate U.S. law, though it suggests leaving final fair-use determinations to the courts. The plan also includes a ratepayer protection pledge to ensure tech firms cover the electricity costs of their data centers.

Reactions are split along partisan lines. House Republican leaders and industry groups like NetChoice welcomed the roadmap, while Senator Marsha Blackburn proposed the more aggressive TRUMP AMERICA AI Act to codify the president's agenda. Conversely, Democratic lawmakers and California Governor Gavin Newsom criticized the framework as a move to gut consumer protections. Representative Don Beyer introduced the GUARDRAILS Act on March 20 to repeal the framework and block the moratorium on state regulations. Separately, the administration canceled all government contracts with Anthropic after CEO Dario Amodei insisted on restrictions preventing the AI Claude from being used for spying or autonomous weapons.