National Policy Framework for AI: Legislative Recommendations (summary)
The seven pillars of AI Governance
https://www.whitehouse.gov/wp-content/uploads/2026/03/03.20.26-National-Policy-Framework-for-Artificial-Intelligence-Legislative-Recommendations.pdf
On March 20, 2026, the White House Office of Science and Technology Policy issued a 4-page document outlining seven areas where the Trump Administration is calling on Congress to act on AI policy. The overarching thrust is to establish federal primacy, promote US AI dominance, and preempt a “patchwork” of state regulations.
The Seven Pillars
I. Protecting Children and Empowering Parents — Age assurance requirements for AI platforms likely accessed by minors; parental controls; affirming existing child privacy laws (COPPA etc.) apply to AI; tying in the already-signed Take It Down Act (deepfakes) as a First Lady initiative.
II. Safeguarding American Communities — AI data center buildout should not raise residential electricity rates (a “Ratepayer Protection Pledge”); streamline permitting for AI infrastructure; combat AI-enabled fraud targeting seniors; support small businesses with grants and tax incentives.
III. Intellectual Property and Creators — Notably, the Administration states it believes AI training on copyrighted material does not violate copyright law, but will let courts resolve it. Congress encouraged to consider collective licensing frameworks and digital replica protections (voice/likeness), with First Amendment carve-outs.
IV. Free Speech and Preventing Censorship — Government should not coerce AI platforms into content moderation based on partisan agendas; Americans should have redress mechanisms against federal censorship of AI platforms.
V. Innovation and AI Dominance — Regulatory sandboxes; open federal datasets in AI-ready formats; explicitly no new federal AI regulatory body — existing sector-specific agencies handle AI in their domains; industry-led standards preferred.
VI. Workforce and Education — AI training integrated into existing apprenticeship and education programs; study workforce disruption trends; expand land-grant institution capacity for AI youth programs.
VII. Federal Preemption of State AI Laws — The most legally consequential section: Congress should preempt state AI laws that impose “undue burdens,” creating one national standard. States retain authority over: general police powers (fraud, consumer protection, child protection), zoning for AI infrastructure, and their own government use of AI. States cannot regulate AI development itself (deemed inherently interstate/foreign policy), unduly burden lawful AI use, or penalize developers for third-party misuse of their models.