Trump AI Framework vs. State Laws: Republican State Legislators Defy White House Preemption Push

Federal framework would preempt all state AI regulations under a 'minimally burdensome national standard'

States would retain authority only over fraud, child protection, zoning, and their own AI use. The framework includes no regulator, liability pathway, or enforcement body. White House OSTP head Michael Kratsios framed it as necessary to give innovators certainty and avoid a regulatory patchwork.

Framework proposes child safety features as congressional suggestions, not mandates

Even the framework's stated goals around child safety and platform accountability carry no legal teeth — they are recommendations rather than binding requirements. The White House has separately stated it has never told a state it cannot enact child safety protections, though this denial did not specifically address the Utah memo.

Framework would shield AI developers from liability for third-party misuse of their models

No independent oversight bodies or liability frameworks are included, meaning AI companies would face significantly reduced legal exposure under the proposed federal standard compared to existing state laws.

White House killed Utah's AI transparency bill with only a brief, unexplained rejection

Rep. Doug Fiefia (R-Utah), a former Google employee, proposed a bill requiring tech companies to disclose how they protect consumers. A one-line White House memo stated the bill was 'unfixable' and 'against the administration's AI agenda,' with no explanation of what that meant. The bill never came to a vote.

Rep. Fiefia directly cited congressional gridlock as justification for state action: 'they not only will not act, they can't act'

Fiefia's framing undermines the White House's implicit premise that a federal framework is forthcoming — if Congress cannot act, states argue preemption would create a regulatory vacuum, not a unified standard.

Republican state senators in Pennsylvania and Texas are advancing their own AI safety bills despite White House opposition

Pennsylvania Sen. Tracy Pennycuick sponsored the SAFECHAT Act, requiring AI chatbot safeguards against self-harm and violence content. Texas Sen. Angela Paxton said states must preserve their ability to pass laws until strong federal legislation exists. Both Republicans expressed skepticism that federal action will arrive in time.

Framework follows a December 2025 executive order targeting state AI laws and potentially linking compliance to federal funding

The Commerce Department list of 'onerous' state AI laws had not yet been published as of prior reporting. David Sacks is the ideological architect of the administration's minimally burdensome posture on AI regulation.

AI development classified as 'inherently interstate,' tied to national security and foreign policy to justify federal primacy

This classification is the legal and political foundation for the preemption argument — framing AI regulation as a federal domain analogous to foreign policy rather than a consumer protection matter states traditionally govern.