Federal vs. State: The AI Preemption Battle Reshaping American Tech Governance

The Executive Order That Changed Everything

On December 11, 2025, President Trump signed an executive order titled “Ensuring a National Policy Framework for Artificial Intelligence” — one of the most aggressive federal interventions in technology governance in decades. The order directs the Department of Justice to establish an AI Litigation Task Force within 30 days, with an explicit mandate to challenge state-level AI regulations in federal court on the grounds that they unconstitutionally burden interstate commerce, conflict with federal regulations, or are otherwise unlawful.

The order does not stop at litigation. It directs the Secretary of Commerce to publish a comprehensive evaluation of existing state AI laws by March 11, 2026, identifying those deemed “onerous” and in conflict with the national policy framework. The review must flag state laws appropriate for referral to the DOJ’s new task force, effectively creating a pipeline from regulatory inventory to legal challenge.

But the executive order’s most consequential provision may be its use of federal funding as leverage. It instructs the Department of Commerce to condition broadband infrastructure funding under the Broadband Equity, Access, and Deployment (BEAD) program on states’ willingness to repeal or modify AI regulations that the federal government deems obstructive. For cash-strapped state governments that have been counting on BEAD grants to close the digital divide, this creates an agonizing choice between AI governance autonomy and desperately needed broadband investment.

The constitutional implications are significant. While the Commerce Clause has long given the federal government authority to regulate interstate commerce, the application of that authority to preempt state AI safety regulations treads into contested territory. Legal scholars point to the 2023 Supreme Court decision in National Pork Producers Council v. Ross, which upheld California’s right to regulate interstate commerce through product standards, as evidence that states retain substantial regulatory authority even over interstate markets.

The State AI Regulation Landscape

To understand why the federal government has taken such an aggressive posture, one must first grasp the complexity of the state-level AI regulatory environment that has emerged in recent years.

In 2025 alone, 46 states introduced over 600 pieces of AI-related legislation, with roughly 145 AI-related bills enacted into law across all 50 states. These laws vary enormously in scope, approach, and stringency. Colorado’s AI Act (SB 24-205), signed by Governor Polis in May 2024 and set to take effect June 30, 2026 after a delay from the original February 2026 date, imposes comprehensive impact assessments and bias testing requirements on deployers and developers of high-risk AI systems — those that make or substantially factor into consequential decisions in employment, education, lending, healthcare, housing, insurance, and legal services.

Illinois’s HB 3773, effective January 1, 2026, amends the state Human Rights Act to make it a civil rights violation for employers to use AI in ways that discriminate against protected classes, even unintentionally. It requires employers to notify employees when AI is used in employment decisions and mandates four-year recordkeeping requirements. California has enacted 24 AI-related laws across its 2024 and 2025 legislative sessions, including the Transparency in Frontier Artificial Intelligence Act (SB 53), signed in September 2025. New York City’s Local Law 144 requires annual bias audits of automated employment decision tools and public disclosure of results.

The compliance burden for companies operating nationally is immediate and substantial. A large employer using AI in hiring must simultaneously comply with Illinois’s disclosure requirements, New York City’s audit mandates, California’s transparency rules, and Colorado’s impact assessment framework, each with different definitions, thresholds, and enforcement mechanisms.

Texas offers an instructive contrast. Its Responsible AI Governance Act (TRAIGA), signed by Governor Abbott on June 22, 2025 and taking effect January 1, 2026, addresses AI discrimination but employs an intent-based liability framework rather than the disparate impact approach used by Colorado and Illinois. Under TRAIGA, an AI system is unlawful only if it was deployed with the intent to discriminate against a protected class. This philosophical divergence means that a single AI hiring tool could be simultaneously lawful in Texas and presumptively problematic in Colorado, depending on how liability is assessed.

The DOJ AI Litigation Task Force

The DOJ AI Litigation Task Force, which became operational by January 10, 2026, represents something unusual in American regulatory history: a federal entity whose primary mission is to challenge state-level technology regulations through litigation.

The Task Force operates under the Attorney General’s authority and has been directed to challenge state AI laws on multiple legal grounds. Its litigation strategy appears to center on three theories. First, it argues that state AI regulations are preempted under the Commerce Clause, particularly where they impose requirements on AI systems used in interstate commerce. Second, it challenges regulations that it claims conflict with First Amendment protections, arguing that AI outputs in certain contexts constitute protected speech. Third, it invokes the Supremacy Clause to argue that state regulations conflict with federal policy as expressed through the executive order and related agency guidance.

The Commerce Department’s March 11 review is designed to feed directly into the Task Force’s litigation pipeline. Regulations flagged as “onerous” will be prioritized for legal challenge. Colorado Attorney General Phil Weiser has already indicated that the state plans to challenge the executive order in court, and early indications suggest the Task Force will focus its initial efforts on Colorado’s AI Act and California’s various AI-related requirements, both because of their comprehensive scope and because successful challenges there would have the greatest chilling effect on other states considering similar legislation.

The executive order also enlists other federal agencies. It directs the Federal Trade Commission to issue a policy statement by March 11, 2026, classifying state-mandated bias mitigation as a potentially deceptive trade practice. The Federal Communications Commission has likewise been directed to work with the DOJ to align with the administration’s AI policies — a move that prompted 23 state attorneys general to file a letter opposing FCC preemption of state AI laws on December 19, 2025.

State Attorneys General Push Back

The federal preemption push has not gone unanswered. Even before the executive order was signed, a bipartisan coalition of 36 state attorneys general sent a letter to Congressional leaders on November 25, 2025, urging them to reject proposals for a federal moratorium that would prohibit states from enacting or enforcing AI laws. The coalition argued that states must remain empowered to apply existing laws and develop new approaches to meet the challenges posed by AI.

The coalition framed the dispute as a fundamental question of federalism. Their letter highlighted real harms from AI, including scams, deepfakes, and inappropriate interactions targeting vulnerable populations like children and seniors, and argued that broad federal preemption would undermine states’ ability to respond quickly and effectively to these emerging risks.

This argument carries particular force because Congress has not passed comprehensive federal AI legislation. A proposed 10-year moratorium on state AI law enforcement was included in the House version of the “One Big Beautiful Bill” in May 2025, which would have preempted existing state AI laws in California, Colorado, New York, Illinois, and Utah, along with over 1,000 pending AI bills. However, the Senate voted 99-1 to remove the moratorium provision, and the bill was signed into law by President Trump on July 4, 2025 without the AI preemption language. The state attorneys general argue that preempting state regulations without providing a federal alternative creates a regulatory vacuum that leaves Americans unprotected.

Following the executive order, opposition intensified. On December 9, 2025, 42 state and territorial attorneys general sent a letter to 13 major AI companies expressing concerns about AI safety issues. The political dynamics are complex — while the opposition skews Democratic, it includes Republican attorneys general and even Republican governors like Ron DeSantis, who posted publicly that “An executive order doesn’t/can’t preempt state legislative action.” This bipartisan element complicates the administration’s efforts to frame opposition as purely partisan.

The BEAD Funding Lever

The conditioning of BEAD program funding on state AI regulatory compliance represents perhaps the most consequential aspect of the executive order — and its most legally vulnerable.

BEAD was established under the Infrastructure Investment and Jobs Act of 2021, with a total allocation of $42.45 billion for broadband deployment in underserved areas. The executive order specifically targets the “non-deployment” portion of BEAD funding — an estimated $21 billion in remaining funds after states meet infrastructure deployment requirements — conditioning eligibility on states rolling back AI regulations deemed onerous.

For rural and underserved communities, the stakes are significant. BEAD funding represents the most substantial broadband investment in American history, and many communities have no alternative path to broadband access. State officials face the prospect of explaining to constituents that their broadband funding has been delayed because the state refused to modify its AI regulation posture.

The legal precedent here is contested. The Supreme Court’s 2012 decision in NFIB v. Sebelius established that the federal government cannot use funding conditions so coercive as to effectively compel state compliance — the Court struck down the Affordable Care Act’s Medicaid expansion enforcement mechanism on those grounds. Legal analysts at Lawfare have argued that the state attorneys general preparing to challenge the BEAD condition “have the better reading of the BEAD statute,” noting that the statute — which is about deploying broadband service and connecting locations — never mentions AI. The major questions doctrine requires clear congressional authorization before an agency asserts power over questions of vast economic and political significance, and federalism-protecting canons require the same before Congress conditions federal funds on state policy.

What Happens at the March 11 Deadline

The March 11, 2026, deadline for the Commerce Department’s regulatory review is the next critical inflection point in this dispute. The review’s contents will determine which state regulations the DOJ Task Force prioritizes for challenge and which states face the most immediate BEAD funding pressure.

For companies operating AI systems nationally, the uncertainty is itself a significant burden. Compliance teams cannot plan effectively when the regulatory ground may shift dramatically within months. Several major technology companies have reportedly paused AI compliance investments pending clarity on which state laws will survive the federal challenge.

The broader implications extend well beyond AI. If the federal government successfully establishes a precedent for preempting state technology regulations through executive action and funding conditions, the framework could be applied to state privacy laws, content moderation requirements, and other technology governance areas where states have acted in the absence of federal legislation. Legal experts like John Bergmayer of Public Knowledge have argued that the administration is “trying to find a way to bypass Congress with these various theories in the executive order” and that the legal theories do not hold up well.

For the moment, the only certainty is uncertainty. The March 11 deadline will clarify the federal government’s specific targets, but the legal battles that follow will likely take years to resolve. In the interim, companies, states, and consumers must navigate a regulatory environment where the rules may change dramatically depending on outcomes in courtrooms and Congress that no one can reliably predict.

Frequently Asked Questions

Sources & Further Reading

• Executive Order: Ensuring a National Policy Framework for Artificial Intelligence — The White House
• Trump Is Trying to Preempt State AI Laws via an Executive Order. It May Not Be Legal — NPR
• The AI Preemption Executive Order’s BEAD Strategy Faces Steep Legal Hurdles — Lawfare
• Bipartisan Coalition of 36 State Attorneys General Opposes Federal Ban on State AI Laws — NAAG
• President Trump Signs Executive Order Challenging State AI Laws — Paul Hastings
• Artificial Intelligence 2025 Legislation — National Conference of State Legislatures