EO: Ensuring a National Policy Framework for AI (State Preemption)
EO: Ensuring a National Policy Framework for AI (State Preemption)
This article covers the December 2025 Trump executive order directing federal agencies to challenge state AI regulations in favor of federal uniformity. The order established a DOJ AI Litigation Task Force and threatened to withhold broadband funding from states with AI regulations, triggering opposition from 36 state attorneys general.
Quick Assessment
| Attribute | Detail |
|---|---|
| Type | Executive Order (United States) |
| Issuing Authority | President Donald Trump |
| Date Signed | December 11, 2025 |
| Primary Goal | Challenge state AI regulations deemed inconsistent with federal policy |
| Key Mechanism | DOJ AI Litigation Task Force; Commerce Department state law evaluation; BEAD funding withholding |
| AI Safety Relevance | Significant — targets state-level AI safety obligations; shifts governance to federal level |
| Status | Enacted; DOJ Task Force established January 9, 2026 |
Overview
This executive order was signed on December 11, 2025, as part of a broader effort to consolidate AI governance at the federal level and prevent state regulations from constraining AI development in the United States. The order proceeds from the premise that inconsistent state-level AI laws create compliance burdens for AI developers and deployers, and that a unified national framework is necessary to preserve American competitiveness in AI.1
The order directs federal agencies to identify state laws and regulations that conflict with federal AI policy, and it establishes enforcement mechanisms — including a DOJ AI Litigation Task Force authorized to challenge state AI laws on interstate commerce, federal preemption, and First Amendment grounds. This represents a significant shift in the regulatory terrain for AI governance and policy, moving authority away from states that had begun enacting their own AI oversight regimes.1
Importantly, the executive order does not itself overturn any state law — only Congress or courts can do that. Its mechanisms are indirect: litigation, federal reporting standards that preempt conflicting state requirements, and financial pressure through withholding of broadband infrastructure funding.1
The order is closely related to broader Trump-era AI policy, including Executive Order 14179 (which revoked the Biden administration's AI executive order) and the subsequent National AI Legislative Framework proposed in March 2026. Critics argue that preempting state-level regulation eliminates existing protections without replacing them with equivalent federal safeguards, while proponents contend that regulatory fragmentation poses genuine obstacles to responsible AI deployment at scale.
Background and Context
By late 2025, a substantial number of U.S. states had enacted or were considering AI-related legislation. These ranged from disclosure requirements for AI-generated content to regulations on algorithmic hiring tools, automated decision-making in consequential domains, and safety testing mandates for frontier AI models. California attracted particular attention: SB 1047, which would have imposed safety obligations on developers of large AI models, passed the California legislature before being vetoed, and California SB 53 enacted more limited transparency and incident-reporting requirements.
This proliferation of state activity reflected both genuine legislative concern about AI harms and the absence of comprehensive federal AI legislation. However, it also produced the multi-jurisdictional compliance landscape that the Trump administration cited as a reason for federal preemption. The administration's position, consistent with broader deregulatory priorities, was that federal uniformity would lower barriers to AI deployment and that safety concerns should be addressed through a single national framework rather than a mosaic of state rules.
The executive order fits within a pattern of Trump-era AI policy that prioritized removing regulatory constraints perceived as hindering innovation. Executive Order 14179, signed in January 2025, had revoked the Biden executive order on AI safety, directed agencies to reduce burdensome AI rules, and signaled skepticism toward precautionary approaches to AI governance.
Key Provisions
The executive order contains four principal directives:1
-
AI Litigation Task Force: The Attorney General was directed to establish a task force within 30 days to challenge state AI laws on interstate commerce, federal preemption, and First Amendment grounds. Attorney General Bondi established this task force on January 9, 2026.
-
State Law Evaluation: The Commerce Secretary must publish an evaluation of state AI laws deemed "onerous" within 90 days, identifying specific laws the administration considers inconsistent with federal policy.
-
Federal Reporting Standard: The FCC was directed to consider establishing a federal AI reporting and disclosure standard that would preempt conflicting state requirements.
-
BEAD Funding Withholding: The order withholds Broadband Equity, Access, and Deployment (BEAD) program funding from states that maintain targeted AI regulations, using federal infrastructure funding as financial leverage.
The order does not itself establish substantive federal AI safety requirements, though it references the concurrent effort to develop such requirements through a national legislative framework.
Relationship to AI Safety
The executive order's implications for AI safety are contested and depend heavily on what federal framework, if any, ultimately fills the space vacated by preempted state laws.
State-level AI legislation had begun to create concrete safety obligations — including requirements for pre-deployment testing, transparency reporting, and incident disclosure — that the preemption order could nullify or chill. The California SB 1047 debate reflected substantive disagreement about whether developers of large frontier models should bear legal responsibility for foreseeable safety failures. By targeting state authority to impose such obligations, the order reduces near-term regulatory pressure on AI developers to invest in safety measures.
Proponents argue that fragmented state regulation is an inefficient vehicle for addressing systemic AI risks, because AI systems are deployed nationally and globally rather than within single state jurisdictions. A coherent federal framework, they contend, could impose more rigorous and consistently enforced safety requirements than a patchwork of state rules — though this outcome depends on the content of the federal framework actually developed.
The order's interaction with AI policy effectiveness debates is significant: it forecloses one set of policy levers (state regulation) while the effectiveness of the replacement (federal framework) remains uncertain. For those concerned about AI existential risk, the removal of state-level safety obligations without demonstrated federal equivalents represents a potential step backward in the near term.
Stakeholder Responses
Opposition
A coalition of 36 state attorneys general formed bipartisan opposition to federal preemption of state AI regulatory authority, representing the most organized governmental resistance to the order.1
The American Civil Liberties Union called the executive order unconstitutional, arguing it exceeded presidential authority over state legislative powers.
State governors in California, Colorado, and New York issued statements defending their states' authority to regulate AI and signaling intent to continue enforcement of existing laws. Congress had previously twice rejected preemption provisions in federal AI legislation, suggesting limited legislative appetite for the approach.
Support
Large technology companies and AI developers generally welcomed the order's deregulatory direction, consistent with industry advocacy against multi-state compliance burdens. The administration framed the order as protecting American AI innovation and competitiveness.
Legal Uncertainties
The executive order's legal authority to preempt state law is contested. Federal preemption typically requires either explicit congressional authorization or a determination that state law conflicts with federal statute. An executive order alone has limited preemptive force against state legislation absent supporting legislation or agency rulemaking.
The order's mechanisms — litigation, federal standards, and funding withholding — represent indirect approaches to preemption rather than direct legal displacement of state laws. The DOJ AI Litigation Task Force can file suits challenging specific state laws, but courts will independently determine whether those laws are actually preempted.
Whether states will continue to pursue AI regulation despite the preemption signal, or whether the order will significantly reduce state legislative activity through its chilling effect, remains an open question.
Key Links
| Source | Link |
|---|---|
| Full Text | White House Presidential Action |
| Legal Analysis | Ropes & Gray Analysis |
Sources
Footnotes
References
1Ensuring a National Policy Framework for Artificial Intelligence (The White House, December 11, 2025)whitehouse.gov·Government▸
This White House executive action establishes a federal preemption framework for AI policy, aiming to eliminate conflicting state-level AI regulations in favor of a unified national approach. It asserts federal supremacy over AI governance to prevent a patchwork of state laws that could obstruct national AI development and deployment priorities. The order reflects the administration's intent to accelerate AI adoption by reducing regulatory fragmentation.