White House AI Framework Calls on Congress to Preempt State AI Laws, Leverages Federal Funding (2026)

Incident Summary

The White House released the “National Policy Framework for Artificial Intelligence” on March 20, 2026. The framework, which still requires congressional action to have preemptive legal force, calls on Congress to preempt state AI laws that “impose undue burdens.”^[1] The framework was prepared by the White House Office of Legislative Affairs with input from Special Advisor David Sacks and OSTP Director Michael Kratsios.^[2]

The framework proposed that states should not regulate AI development, should not penalize developers for third-party misuse, and should not “unduly burden Americans’ use of AI for activity that would be lawful if performed without AI.” Three carve-outs were preserved for state authority: traditional police powers (child safety, fraud prevention, consumer protection), zoning authority over AI data center siting, and state governments’ own procurement and use of AI.^[1]

The December 2025 executive order that preceded the framework established enforcement mechanisms: a DOJ AI Litigation Task Force to challenge state laws in federal court, a Commerce Department review to identify “onerous” state AI laws, and a provision making states with such laws ineligible for remaining funds under the $42 billion Broadband Equity, Access, and Deployment (BEAD) Program.^[2]

The Colorado AI Act was explicitly named as a problematic example of state legislation. In 2025, more than 1,000 AI-related bills were introduced across all U.S. states and territories, reflecting the state-level regulatory activity the framework seeks to curtail.^[2]

AI Progress, an industry coalition whose members include Amazon, Anthropic, Google, Meta, Microsoft, Midjourney, and OpenAI, publicly welcomed the framework. A bipartisan GUARDRAILS Act (Guaranteeing and Upholding Americans’ Right to Decide Responsible AI Laws and Standards Act) was introduced in both chambers to repeal the executive order, led by Senator Brian Schatz (D-HI) with six co-sponsors in the Senate. The bill had not been enacted at the time of writing.^[2]

Legal experts noted that the executive order alone likely cannot preempt state laws without congressional legislation, as federal preemption typically requires enacted statute.^[3][4]

Key Facts

• 1,000+ state AI bills in 2025: The framework targets a broad wave of state-level AI regulation^[2]
• Colorado AI Act named: Explicitly identified as problematic; scheduled to take effect June 2026^[2]
• DOJ task force: Established to challenge state AI laws in federal court^[2]
• BEAD funding leverage: States with “onerous” AI laws become ineligible for portions of the $42 billion broadband program^[2]
• Industry coalition support: AI Progress (Amazon, Anthropic, Google, Meta, Microsoft, OpenAI) publicly endorsed the framework^[2]
• Bipartisan opposition: GUARDRAILS Act introduced in both chambers to repeal the executive order^[2]
• State resistance: Governors of California, Colorado, and New York stated the order would not deter their legislation^[2]

Threat Patterns Involved

Primary: Safety Governance Override — State AI safety legislation represents existing governance infrastructure built through democratic processes; the framework seeks to override or constrain these laws via federal preemption mechanisms, DOJ challenges, and funding leverage. Several targeted laws, including the Colorado AI Act, were already enacted or scheduled to take effect.

Secondary: Accumulative Risk & Trust Erosion — The framework removes safety governance at the state level without establishing equivalent federal protections. The “light-touch” approach using existing federal agencies may create a governance gap during the transition period.

Secondary: Power & Data Concentration — The preemption framework was welcomed by a coalition of the largest AI companies (Amazon, Google, Meta, Microsoft, OpenAI), who benefit from a single, minimally burdensome federal standard rather than compliance with diverse state requirements.

Significance

1. Democratic governance vs. executive action: The framework uses executive mechanisms (DOJ task force, funding leverage) to override state legislation enacted through democratic processes, raising constitutional questions about federal authority over state AI regulation.
2. Funding leverage as a regulatory tool: Conditioning $42 billion in broadband funding on AI policy compliance creates a financial penalty for states that regulate AI, a mechanism that goes beyond traditional preemption.
3. Governance gap risk: Eliminating state AI laws without establishing equivalent federal protections creates a period where AI systems operate under reduced oversight. The “light-touch” approach assumes existing laws adequately cover AI-specific risks, an assumption that state legislatures have evidently rejected by passing AI-specific legislation.
4. Industry-government alignment: The explicit endorsement by AI Progress members, who are both subjects of state regulation and beneficiaries of preemption, has been cited by critics as indicative of potential regulatory capture in AI policy.