Executive Order Targets State AI Laws With Litigation and Funding Threats

The most consequential AI rules in the United States are increasingly written in state capitols, not Washington. Congress has failed to deliver a single, durable federal framework that tells developers, deployers, and regulators exactly what “safe,” “fair,” and “accountable” look like at scale. On Dec. 11, the White House responded with an executive order that treats state AI laws less like policy experiments and more like obstacles. Ensuring a National Policy Framework for Artificial Intelligence sets up a federal strategy built around litigation, agency positioning, and funding leverage. The result is not a federal safety code, but a federal effort to stop states from building one.

Choosing Negative Regulation Over Federal Standards

If the administration’s theory holds, 2026 becomes a preemption year. Instead of companies harmonizing state-by-state obligations, they watch courts and agencies decide how much of that state law can survive a White House push for a single national approach. If the theory fails, states will legislate faster, emboldened by a public fight that frames them as the only institutions willing to write enforceable rules.

The easiest way to understand the executive order is to focus on what it omits. No federal licensing scheme appears for high-risk AI. No national duty of care is created for algorithmic discrimination. No mandatory audit, incident reporting, or “safety case” regime is established for advanced models. The document instead declares a policy objective of a “minimally burdensome” national framework and casts “excessive” state regulation as a threat to innovation and competitiveness.

The White House’s fact sheet makes the posture even clearer. The message is that the national interest is best served by preventing “a patchwork” of state AI regulations, and that the near-term federal job is to challenge state laws that are “unconstitutional, preempted, or otherwise unlawful.” In other words, the administration is not saying, “we will regulate AI better than the states.” The message is, “we will regulate less, and we will use federal tools to keep states from regulating more.”

That is why the executive order reads like negative regulation. National uniformity is pursued by limiting the number of enforceable obligations, not by replacing those obligations with a robust national standard. In the short run, that is a litigation plan. In the long run, it is a legislative ask, paired with an attempt to freeze the field before Congress acts.

Three Enforcement Mechanisms Take Shape

The executive order is operational, not symbolic. It sets deadlines, creates a task force, and directs agencies to build a record that can support future lawsuits and preemption positions. The core machine has three moving parts: the Justice Department, the Commerce Department, and federal funding leverage.

First, within 30 days, the attorney general is directed to establish an “AI Litigation Task Force” with the stated purpose of challenging state AI laws inconsistent with the order’s policy. The order points the task force toward theories that are familiar in structure, even if new in application: dormant Commerce Clause arguments about regulating beyond state borders, preemption arguments tied to existing federal regimes, and broader claims that certain state provisions are unlawful.

Second, within 90 days, the Commerce Department must publish an evaluation identifying state AI laws it views as “onerous,” including laws that allegedly compel model outputs to change, or require disclosures and reporting that the administration characterizes as constitutionally suspect. The evaluation is not just a report card. The order contemplates referrals from Commerce to DOJ, turning an agency assessment into a litigation pipeline.

Third, the order ties this to money. It instructs Commerce to issue a policy notice conditioning eligibility for certain remaining nondeployment funding under NTIA’s Broadband Equity, Access, and Deployment (BEAD) program. States with AI laws the evaluation labels “onerous” are positioned as ineligible for those categories, to the maximum extent permitted by law. This is the sharpest leverage in the order because it attempts to translate federal dissatisfaction with state technology regulation into immediate fiscal consequences.

The order also points beyond broadband, directing agencies to review discretionary grant programs and consider conditions that would require states to refrain from enforcing laws deemed inconsistent with national AI policy, including agreements not to enforce during a grant’s performance period. That is the kind of provision that can look small on paper and feel enormous in practice, because it pressures states to trade enforcement authority for federal dollars.

Finally, it nudges agencies toward preemption postures. The order directs the Federal Communications Commission to initiate a proceeding to consider a federal AI reporting and disclosure standard that could preempt conflicting state laws, and it directs the Federal Trade Commission to issue a policy statement on how the FTC Act applies to AI models, including when state laws that require changes to “truthful outputs” could be treated as requiring deception. The executive order itself does not preempt. It tries to build a federal record and a set of federal positions that can support preemption claims later.

Executive Orders Cannot Erase State Statutes

Preemption in the U.S. system typically flows from Congress. A statute can expressly preempt state law, or it can create a federal framework so comprehensive that it implies field preemption, or it can conflict with state law in a way that triggers conflict preemption. Executive orders are different. They direct the executive branch. They can set enforcement priorities. They can shape agency rulemaking agendas. But they do not erase state statutes by fiat.

The White House seems to recognize that constraint by directing officials to produce a legislative recommendation for a national AI framework that would preempt conflicting state laws. It also includes the standard caveat that the order does not create any right or benefit enforceable at law against the United States. That disclaimer will not stop lawsuits. It does frame what is actually happening: the administration is using executive power to pressure states and position federal agencies, while asking Congress to provide the durable preemption vehicle later.

That is why 2026 is likely to feel less like a clean federal takeover and more like a contested transition. The path to actual preemption runs through courts, statutes, and agency actions that must rest on real congressional authorization.

Four Constitutional Battlegrounds Frame Federal Strategy

The executive order points to several legal battlegrounds, but four are likely to shape the early phase: conditional spending, the dormant Commerce Clause, First Amendment framing, and agency preemption arguments.

The broadband funding component is an invitation to Spending Clause litigation. The central questions will be familiar: did Congress authorize the condition, is the condition related to the purpose of the program, and does the condition cross the line into coercion. Reuters has reported that legal experts see substantial hurdles for the administration, including whether Congress intended broadband funding to be used as leverage against state AI regulation, and whether courts will accept the connection between these two policy domains.

The dormant Commerce Clause theory is the other big lever, and it is explicitly referenced in the executive order’s policy framing about state regulation “beyond State borders.” The legal difficulty is that a state law can be costly without being unconstitutional. Many dormant Commerce Clause cases turn on whether a statute discriminates against out-of-state firms or unduly burdens interstate commerce in a way courts are willing to police. The administration’s approach suggests it plans to argue that certain AI laws, especially those dictating how model outputs must behave or how disclosures must be made nationwide, create extraterritorial effects that courts should treat as impermissible.

The First Amendment framing is woven into the order’s instructions to identify laws that compel disclosures and reporting in constitutionally suspect ways. If the Commerce evaluation labels certain state disclosure regimes as compelled speech, and if DOJ sues, courts will be asked to decide whether state AI transparency obligations are ordinary consumer protection measures or unconstitutional mandates. That question has higher stakes than a single statute, because it could shape how far states can go on transparency and documentation when Congress does not act.

Agency preemption is the slowest and most technically complex battlefield. An FCC proceeding can only preempt if the FCC has the statutory authority to regulate in the way it proposes and if it builds a record strong enough to survive administrative law review. The FTC policy statement is even more novel. It attempts to align the FTC Act’s anti-deception posture with a claim that some state “truthful output” mandates require deception. If agencies adopt these theories, the litigation risk shifts from “states sue the administration” to “companies argue preemption defensively,” and that is when compliance teams will want hard answers that courts may take years to supply.

State Statutes Already on the Books

The executive order does not need to list 50 statutes to make its point. It only needs one or two examples that represent what states have been trying to do. Colorado, California, Illinois, Utah, and Tennessee illustrate the basic problem: states are building AI governance regimes because the federal government has not, and the federal government is now trying to halt those regimes before they harden into national norms.

Colorado’s SB 24-205, Consumer Protections for Artificial Intelligence, is one of the clearest attempts to legislate around algorithmic discrimination in high-stakes contexts. It imposes obligations on developers and deployers of “high-risk” AI systems and requires reasonable care to protect consumers from known or reasonably foreseeable risks of algorithmic discrimination, with key obligations taking effect Feb. 1, 2026. The executive order’s critique of a “new Colorado law” banning algorithmic discrimination signals that statutes like SB 24-205 are likely to be treated as test cases, especially if federal officials frame their obligations as compelled content shifts or burdensome disclosure requirements.

California’s SB 53, signed by Gov. Gavin Newsom on Sept. 29, 2025, reflects a different state instinct: focus on advanced systems, transparency, and public trust. The governor’s office described the law as advancing an AI strategy while putting “commonsense” guardrails in place. For companies building nationwide programs, California’s approach matters because it can become an anchor requirement, the kind of rule that national product teams build around. For the White House, it looks like another brick in the wall of a state-led governance regime that the administration wants to prevent from becoming de facto national policy.

Illinois amended the Illinois Human Rights Act in 2024 to prohibit AI use that causes discriminatory effects in employment decisions, with obligations taking effect Jan. 1, 2026. Utah enacted the Artificial Intelligence Policy Act in 2024, which requires disclosure when generative AI interacts with consumers and creates an Office of AI Policy to oversee regulatory mitigation agreements. Tennessee’s ELVIS Act protects voice and likeness against AI impersonation, targeting unauthorized use in creative works. While the White House moves to freeze these specific AI statutes, state regulators maintain a secondary, more durable set of tools that remain insulated from federal preemption.

State Enforcement Survives Without AI Statutes

An aggressive federal strategy still leaves the state enforcement ecosystem intact. States do not need an “AI Act” to regulate AI, because consumer protection, privacy, labor rules, professional licensing, and procurement authority can all shape how systems are built and deployed. The executive order acknowledges that reality indirectly by targeting “state AI laws” while carving out categories that are politically difficult to preempt.

Start with consumer protection. State unfair and deceptive practices statutes remain the workhorse enforcement tool when AI marketing exaggerates capabilities, disclosures are incomplete, or automated decisions cause consumer harm. A preemption strategy that targets AI-specific state statutes does not automatically neutralize general UDAP authority. That is one reason the White House also directs FTC positioning, because consumer deception is the lane where federal and state authority often overlap rather than replace one another.

Privacy and biometric laws are another durable state lane. State privacy regimes govern data collection, retention, sharing, and notice, and those duties attach to AI systems because models ingest and transform personal data. Biometric rules shape the legality of facial recognition, voiceprints, and other sensitive identifiers. These are not niche issues. They are the data backbone of modern AI deployment.

Professional licensing remains state-dominated. States regulate the practice of medicine, law, mental health services, insurance, and other licensed activity. AI tools that influence diagnosis, benefits decisions, legal advice, or hiring can trigger state oversight regardless of whether a state has enacted a dedicated AI statute.

Procurement and government use are also key, and here the executive order’s own posture matters. It instructs officials not to propose preempting otherwise lawful state laws that govern state government procurement and use of AI. That means state and local contracting requirements, including vendor documentation, testing expectations, and usage limits, can continue to spread even if broad state AI statutes are chilled. This is one reason the National Conference of State Legislatures framed its response as a call for collaboration rather than preemption, with Illinois State Rep. Marcus Evans Jr. and Montana Sen. Barry Usher stating: “The best path forward is partnership, not preemption.”

The practical takeaway is that preemption pressure may narrow certain state AI mandates, but the expectation remains that AI systems must comply with state consumer, privacy, and sector laws. Preemption pressure mainly shifts which parts of the state toolkit get used most aggressively.

Global Implications of Domestic Fragmentation

The executive order’s attempt to freeze state-level AI governance while Congress remains gridlocked carries consequences beyond U.S. borders. The European Union’s AI Act entered force in August 2024 and is establishing the world’s first comprehensive risk-based regulatory framework, with full implementation by August 2026 for most provisions. The EU framework creates clear obligations for high-risk AI systems, transparency requirements for general-purpose models, and enforcement mechanisms already operational.

If the United States enters a prolonged period of federal-state litigation while the EU continues building out technical standards, conformity assessments, and enforcement coordination, American influence on emerging global AI norms may weaken at a critical juncture. The “Brussels Effect,” where EU regulations become de facto global standards due to market access requirements, has already shaped data privacy through GDPR. A fragmented U.S. approach during the formative years of international AI governance risks ceding leadership on questions of algorithmic accountability, transparency obligations, and cross-border data use to jurisdictions with functioning regulatory frameworks. The executive order’s focus on preventing state action, rather than building federal capacity, leaves the United States without a credible alternative model to offer international standard-setting bodies at the moment those standards are being written.

Industry Response Splits Along Predictable Lines

The executive order represents a win for tech companies that have argued a patchwork of state laws hinders U.S. competition with China. Yet industry response has been more divided than the White House narrative suggests. Large technology firms have lobbied for federal uniformity, while some startup executives and legal advisors warn that the order creates legal limbo rather than clarity.

TechCrunch reported that Andrew Gamino-Cheong, CTO of AI governance company Trustible, argued the order will backfire on innovation: “Big Tech and the big AI startups have the funds to hire lawyers to help them figure out what to do, or they can simply hedge their bets. The uncertainty does hurt startups the most.”

California Governor Gavin Newsom responded by framing the order as “advanc[ing] corruption, not innovation,” noting that California is “the fourth-largest economy in the world, the birthplace of tech, and the top pipeline for tech talent.” Brad Carson, president of Americans for Responsible Innovation, stated that the order “directly attacks the state-passed safeguards that we’ve seen vocal public support for over the past year, all without any replacement at the federal level.”

Even within the Republican Party, the order has faced pushback. NPR reported that Utah Gov. Spencer Cox posted on social media that he preferred an alternative executive order that did not include barring state laws, writing: “States must help protect children and families while America accelerates its leadership in AI.” Florida Gov. Ron DeSantis similarly expressed opposition, arguing that executive orders cannot preempt state legislative action.

Morgan Reed, president of The App Association, urged Congress to enact “a comprehensive, targeted, and risk-based national AI framework,” adding that “a lengthy court fight over the constitutionality of an Executive Order isn’t any better” than state-by-state rules.

Guidance for Multi-State Deployers

For companies operating across state lines, the temptation will be to treat the executive order as a reason to pause compliance investments. That is the wrong instinct. In a transition period, legal risk does not disappear. It moves into discovery, enforcement discretion, and contract allocation.

The safest posture is to build governance that survives either outcome. If states win, companies need an operational ability to document testing, monitor performance, and show reasonable controls for high-stakes uses. If the federal strategy succeeds, companies still need defensible records because consumer protection and privacy claims do not vanish, and because federal agencies may shift their enforcement posture quickly once they have committed to a national policy theory.

This is also a contracting story. Vendor agreements should anticipate changing legal obligations, including disclosure demands, audit requests, and incident reporting expectations. Customer-facing commitments should avoid sweeping promises about accuracy, bias, or compliance, and should reflect the difference between product design intent and real-world performance. The core discipline is consistency: what a company says publicly, what it writes in contracts, and what it documents internally should match. If they do not, litigation makes the mismatch the plot.

Finally, deployers should plan for procurement divergence. State and local entities can keep demanding documentation and controls even if broader state AI statutes are in court. A company that can answer procurement diligence questions clearly tends to be the company that can survive the first wave of regulatory inquiry intact, too.

What to Watch as 2026 Begins

The executive order sets a short fuse. The Justice Department task force is a 30-day item, while the Commerce evaluation, the BEAD-related policy notice, and the FTC policy statement have 90 days. The FCC proceeding is tied to the Commerce identification process and is designed to move after Commerce publishes its work. That means the first quarter of 2026 is likely to produce the first concrete signals, not because the courts will decide quickly, but because agencies will begin to publish positions that shape how litigation is framed.

The earliest lawsuits are likely to target the funding lever and any attempt to condition federal grants on state non-enforcement commitments. Those disputes will determine whether the administration’s strategy is mostly rhetorical, or whether it can be operationalized in a way that forces states to choose between federal dollars and state regulation.

Meanwhile, the political fight will run in parallel. State lawmakers have already described the approach as federal overreach, while the White House frames it as necessary to prevent a fragmented regulatory environment. National reporting has emphasized that the order faces both political and legal hurdles, and that courts may take a hard look at how far executive power can go when Congress has not enacted a comprehensive AI statute. That tension is the story of 2026: a national AI strategy emerging first as a courtroom plan.

Sources

• The Bond Buyer: “States push back on AI executive order,” by Scott Sowers (Dec. 12, 2025)
• Colorado General Assembly: “SB24-205 Consumer Protections for Artificial Intelligence” (2024)
• Governor of California: “Governor Newsom signs SB 53, advancing California’s world-leading artificial intelligence industry” (Sept. 29, 2025)
• Governor of California: “Trump’s AI executive order advances corruption, not innovation” (Dec. 11, 2025)
• The Guardian: “Trump signs executive order aimed at preventing states from regulating AI,” by Nick Robins-Early and Dara Kerr (Dec. 12, 2025)
• Hinshaw Law: “Strategic Artificial Intelligence Planning Alert: A State and Federal Regulatory Roadmap for 2025 Compliance”
• Investing.com (Reuters): “Trump’s AI order faces political and legal hurdles” by Jody Godoy (Dec. 12, 2025)
• Mallinson, D. J., Azevedo, L., Best, E., & Robles, P. (2025). Artificial intelligence policy, the Trump Administration, and federalism. Administrative Theory & Praxis, 47(3), 202–210. https://doi.org/10.1080/10841806.2025.2537601
• Mayer Brown: “Illinois Passes Artificial Intelligence (AI) Law Regulating Employment Use Cases,” by Arsen Kourinian, Kyla Miller and Charles E. Harris II (Dec. 4, 2024)
• National Telecommunications and Information Administration: “Broadband Equity, Access, and Deployment Program (BEAD)”
• NPR: “Trump is trying to preempt state AI laws via an executive order. It may not be legal,” by Huo Jingnan (Dec. 11, 2025)
• SSRN: “The U.S. Approach to AI Regulation: Federal Laws, Policies, and Strategies Explained,” by Tatevik Davtyan (Oct. 17, 2024)
• TechCrunch: “Trump’s AI executive order promises ‘one rulebook’—startups may get legal limbo instead” by Rebecca Bellan (Dec. 12, 2025)
• The White House: “Ensuring a National Policy Framework for Artificial Intelligence” (Executive Order, Dec. 11, 2025)
• The White House: “Fact Sheet: President Donald J. Trump Ensures a National Policy Framework for Artificial Intelligence” (Dec. 11, 2025)

This article was prepared for educational and informational purposes only. It does not constitute legal advice and should not be relied upon as such. All cases, regulations, and sources cited are publicly available through court filings and reputable media outlets. Readers should consult professional counsel for specific legal or compliance questions related to AI use.

See also: Global AI Regulation Is Becoming the New Baseline for U.S. Legal Risk