New Executive Order Plants the Seeds of a Federal Framework for Artificial Intelligence Regulation

By Derek Flint

The December 11, 2025 Executive Order on “Ensuring a National Policy Framework for Artificial Intelligence”¹ seeks to establish a unified federal approach to AI by mobilizing federal agencies to challenge state laws, condition federal funds on state regulatory choices, and pursue agency actions that would preempt conflicting state rules. The order signals an assertion of federal primacy over AI policy, but it does so largely through executive action and agency processes rather than through legislation. While the order has the laudable aim of ending “patchwork” state regulation in AI, it may invite legal challenges under several legal theories, including that the type of sweeping preemption it seeks to implement generally requires congressional authorization.

The order’s stated purpose is to maintain U.S. leadership and dominance in AI by preventing “excessive State regulation,” criticizing a “patchwork” of state laws and policies — particularly those the Administration views as mandating “ideological bias” or extraterritorial application — and asserting the need for a “minimally burdensome national standard.” It expressly identifies certain state measures — such as “algorithmic discrimination” requirements that are alleged to force “false results” — as problematic and targets such laws for federal challenge.

To effectuate this policy, the order directs a suite of concrete actions:

The order includes standard caveats that must be implemented consistent with law and that do not create enforceable rights. It sets a timeline across multiple agencies to begin curbing, conditioning, or preempting state AI rules in advance of congressional action.

There are a number of key takeaways from this Executive Order — both for states and AI stakeholders. First, the Administration is attempting to implement a national standard for AI primarily through executive direction and agency action, not through a congressionally enacted statute. The order frames “federal dominance” in AI policy as urgent, and positions state laws as the principal impediment to innovation, competitiveness, and national security. This framing is offered to explain the order’s call for rapid federal challenges to state AI rules, conditional funding mechanisms to deter states from enacting or enforcing them, and agency-led preemption strategies.

Second, the DOJ, Commerce, FCC, and FTC are the principal instruments. DOJ will litigate against states; Commerce will catalog and label “onerous” state AI laws; FCC is asked to consider a preemptive reporting/disclosure regime; and FTC is asked to position the FTC Act as a vehicle to nullify certain state mandates on model outputs. This multi-agency approach is designed to allow immediate examination of AI legislation while a longer-term federal statutory framework is developed.

Third, the order links federal funding to state compliance. By tying BEAD allocations and other discretionary grants to states’ willingness to refrain from or suspend enforcement of identified AI laws, the Administration is attempting to reshape the state regulatory environment through conditions, not just litigation. This is likely to be among the most rapidly impactful levers — and among the most litigated.

Fourth, the order explicitly targets state laws that force “alterations to the truthful outputs” of AI models and frames such laws as potentially violating the First Amendment or as preempted by federal prohibitions on deception. The FTC is positioned to argue that state requirements to tweak outputs for fairness or anti-discrimination may compel deception and therefore conflict with federal law. This may present a federal-state dispute over the boundary between consumer protection, anti-discrimination norms, and expressive or technical model behavior.

Fifth, although the order calls for future legislation, its immediate emphasis is on executive and administrative means of preempting or precluding enforcement of state AI rules. This may result in tension among those impacted, as the resulting framework is being implemented before Congress acts.

As with many Executive Orders in important areas of the law, legal challenges to this order are likely. Some of the challenges we foresee include the following.

Separation of Powers and Preemption by Executive Order. Courts often expect clear congressional statements to preempt state law, particularly in domains of traditional state competence. Without that clarity, attempts to bootstrap preemption into agency policy or reporting regimes may be viewed as exceeding executive authority. This tension is compounded by the order’s call for legislation, which arguably acknowledges that Congress — not the Executive — should ultimately establish the uniform national framework.

Statutory Authority and Major Questions Concerns for Agency Preemption. The FCC’s contemplated “Federal reporting and disclosure standard for AI models that preempts conflicting State laws” invites threshold questions about statutory authority to regulate “AI models” as such, the scope of any ancillary jurisdiction, and whether Congress spoke clearly enough to authorize preemptive AI regulation under communications statutes. Similar questions arise for the FTC policy statement invoking the FTC Act to preempt state laws that require altering “truthful outputs,” especially where the FTC Act has long coexisted with state unfair-trade-practice regimes. Without clear textual grounding, these moves could be challenged as exceeding statutory authority or as raising major policy questions that require explicit congressional direction.

Spending Clause and the Constitutionality of Conditions. The order directs Commerce to make states with “onerous” AI laws ineligible for certain BEAD funds “to the maximum extent allowed by Federal law,” and urges agencies to condition discretionary grants on states agreeing not to enact or enforce conflicting AI laws during the grant period. Litigants may argue that these conditions are unduly coercive, insufficiently related to the federal program’s purpose, and imposed after-the-fact on funding streams that states reasonably rely upon — raising potential Spending Clause and unconstitutional conditions issues. These conditions may also be challenged as attempts to commandeer state legislative processes or executive enforcement choices.

Anti-Commandeering and State Sovereignty. Relatedly, conditioning federal funds on a state’s agreement not to enact certain laws, or not to enforce laws that have already been enacted during federal funding periods, could be argued as federal overreach into core state functions. While the order states that it intends to stay within “the maximum extent allowed by Federal law,” challengers may argue that leveraging federal funds to suspend state law enforcement crosses the line from inducement into practical coercion or commandeering of state policy choices.

First Amendment and Viewpoint Discrimination Arguments. The order highlights state requirements that force changes to “truthful outputs” and flags concerns about compelled speech or disclosure. States and their allies may counter that many AI-output and transparency rules are content-neutral consumer protection or anti-discrimination measures, not compelled falsehoods, and that the federal move to characterize them as “deception” rewrites traditional consumer protection frameworks. Relatedly, if federal policy effectively requires “truthful outputs” as the governing standard, line-drawing around “truth,” safety mitigation, and bias correction may become central — and contested.

The order is a signal of federal intent to curb divergent state AI laws through litigation, funding levers, and agency preemption theories. Its immediate impact may be to chill or constrain state regulatory initiatives and to launch significant federal-state litigation. However, because the order relies primarily on executive and administrative tools rather than a congressional mandate, it may face legal headwinds across separation of powers, Spending Clause, administrative authority, and federalism grounds. Businesses that are impacted by state-level AI regulation should monitor both the implementation of this order and any challenges that arise.

**Any opinions expressed are the author’s and not necessarily those of the firm or his colleagues.