The White House has issued an executive order aimed at pre-empting state regulation of artificial intelligence, directing federal agencies to challenge state laws in court and conditioning some funding on states stepping back from rules the Administration deems “onerous.”

Signed by President Donald Trump on 11 December (US time), the order sets a national policy of “minimally burdensome” AI regulation and frames state-by-state laws as a patchwork that burdens startups, risks interstate commerce conflicts, and in some cases “requires AI models to alter their truthful outputs.” It cites Colorado’s new “algorithmic discrimination” law as an example of a measure that could force differential outputs to avoid disparate impact.

Key steps ordered:

• Justice Department litigation: Within 30 days, the Attorney General must create an AI Litigation Task Force to challenge state AI laws viewed as inconsistent with the order’s policy, including on Commerce Clause, preemption, or First Amendment grounds.
• Mapping state laws: Within 90 days, the Secretary of Commerce must publish an evaluation of state AI statutes, flagging laws to be referred for challenge and identifying any that align with the federal policy.
• Funding leverage: Within 90 days, the Commerce Department’s NTIA must issue conditions for remaining Broadband Equity, Access and Deployment (BEAD) funds that render states with “onerous” AI laws ineligible for non-deployment money, to the extent allowed by law. Other agencies are told to review discretionary grants and potentially condition awards on states not enacting or enforcing conflicting AI rules during a funding period.
• Federal standards: The FCC is instructed to consider a federal AI reporting and disclosure standard that would pre-empt conflicting state requirements. Separately, the FTC must issue a policy statement setting out when state laws that require altering “truthful outputs” are pre-empted by the federal ban on unfair or deceptive practices.
• Legislation: The White House will draft recommendations for a federal AI framework that pre-empts conflicting state laws, with carve-outs for child safety protections, state government procurement and use of AI, and most state rules on compute and data centre infrastructure.

The order says its goal is to “sustain and enhance” US AI leadership, arguing that state rules risk embedding “ideological bias” in models and chilling innovation. It follows an earlier directive (Executive Order 14179) that revoked prior constraints and instructed agencies to clear regulatory barriers to AI adoption.

What this means:

• For states: Expect federal lawsuits and potential funding consequences, especially where laws mandate transparency or anti-discrimination controls the Administration says compel model behaviour. The legal fights would likely turn on federal preemption, the Commerce Clause and the First Amendment.
• For companies: A single national baseline would simplify compliance inside the US. The coming FCC and FTC moves point to federal reporting/disclosure rules and limits on state mandates that change outputs.
• For global providers: The US is signalling a lighter-touch, pro-innovation stance that may diverge from the EU’s AI Act and other jurisdictional approaches. Firms operating across markets may face sharper differences between US and European compliance.

What’s next:

• DOJ’s task force must be set up within 30 days.
• Commerce’s assessment of state laws, and NTIA’s BEAD funding criteria, are due within 90 days.
• The FCC proceeding on disclosure and the FTC policy statement on preemption are also due on a similar timeline.
• The Administration will send Congress a proposal for a uniform federal AI framework with specified carve-outs.

The order states it does not create enforceable rights and must be implemented consistent with existing law and available appropriations. It places particular emphasis on pre-empting state rules that, in the Administration’s view, force models to produce non-truthful or differential outputs, and on preventing states from projecting their AI laws beyond their borders.