A constitutional collision is brewing over who gets to write the rules for artificial intelligence in America. The Trump administration is waging an aggressive campaign to centralize AI governance at the federal level, deploying executive orders, a Department of Justice litigation task force, and threats to withhold billions in infrastructure funding from states that refuse to fall in line. But states are not backing down. With more than 45 states having introduced AI-related legislation since 2023 and major laws already taking effect in California, Colorado, Texas, and Illinois, the battle over AI regulation has become the most consequential federalism fight in technology policy in a generation.
The Federal Power Play
The administration's push began in earnest on December 11, 2025, when President Trump signed an executive order titled "Ensuring a National Policy Framework for Artificial Intelligence." The order declares that U.S. AI dominance depends on a "minimally burdensome national policy framework" and warns that state-level regulation threatens to "stymie innovation." It directs federal agencies to identify state AI laws deemed inconsistent with national policy, establishes a DOJ AI Litigation Task Force to challenge those laws in court, and authorizes the Commerce Department to withhold up to $42 billion in Broadband Equity, Access, and Deployment Program funding from states with regulations the administration considers "onerous."
The White House escalated on March 20, 2026, releasing a four-page National Policy Framework for Artificial Intelligence that amounts to a legislative blueprint for Congress. The framework calls for broad federal preemption of state AI laws while organizing federal AI policy around six objectives: protecting children online, managing energy costs, safeguarding intellectual property, preventing AI-driven censorship, promoting innovation, and developing an AI-ready workforce. Critically, the framework asks Congress to preclude states from regulating AI model development or imposing liability on developers for unlawful conduct by third parties using their systems.
"If the Trump Administration tries to enforce this ridiculous order, we will see them in court," California State Senator Scott Wiener, author of the state's AI safety bill, said in a statement after the December executive order was signed.
States Dig In
The state response has been anything but compliant. Colorado's AI Act -- believed to be the first state law in the nation addressing algorithmic discrimination -- went into effect on February 1, 2026, with full enforcement beginning June 30. California's AI Transparency Act and Texas's Responsible AI Governance Act both took effect on January 1, 2026. Illinois now requires employers to notify job candidates and obtain consent before AI evaluates video interviews. New York City continues to enforce Local Law 144, mandating bias audits for automated hiring tools.
Colorado Attorney General Phil Weiser has said the state plans to challenge the executive order in court. A bipartisan coalition of 36 state attorneys general has issued formal opposition to any federal ban on state AI laws, and more than 280 state lawmakers from both parties signed a letter urging Congress to reject preemption legislation.
"I agree on not overregulating," South Carolina State Representative Brandon Guffey, a Republican, told Stateline, "but I don't believe the federal government has the right to take away my right to protect my constituents if there's an issue with AI."
The legal arguments cut both ways. While the administration's AI advisor David Sacks has suggested the federal government can override state laws under the Commerce Clause, critics point to the 2023 Supreme Court decision in National Pork Producers Council v. Ross, which upheld California's right to restrict interstate commerce. John Bergmayer, legal director of the nonprofit Public Knowledge, told NPR that states regulate interstate commerce "all the time" and that the Supreme Court has recently affirmed their authority to do so.
Congress Stalls on Preemption
The White House framework envisions Congress as the vehicle for codifying federal preemption, but early legislative efforts have met fierce resistance. The Senate voted 99-1 to strip a provision from the One Big Beautiful Bill Act that would have preempted state AI regulations for ten years. A similar last-minute attempt to insert preemption language into the National Defense Authorization Act sparked outcry from lawmakers across the ideological spectrum, including Republican Representatives Chip Roy and Thomas Massie and Democratic Senators Ed Markey and Elizabeth Warren.
Senator Marsha Blackburn has introduced the TRUMP AMERICA AI Act, the most ambitious congressional attempt to codify the executive order into comprehensive legislation with federal preemption provisions. But the bill faces an uncertain path. As Morgan Lewis noted in its analysis, "broad preemption will be a difficult sell, in part because many lawmakers remain skeptical of sweeping federal overrides of state AI regulation." Federal AI legislation would require Democratic support to advance, making the political calculus even more complex heading into midterm elections.
Why This Matters
For AI companies, the regulatory uncertainty is itself the problem. Firms operating across multiple states currently face a patchwork of compliance requirements -- bias audits in New York, transparency disclosures in California, algorithmic discrimination rules in Colorado, consent requirements in Illinois. A unified federal standard could dramatically reduce compliance costs, particularly for startups that lack the legal resources to navigate fifty different regulatory regimes.
But consumer advocates warn that federal preemption could strip away protections that states have built precisely because Congress has failed to act. The FTC has been the most active federal enforcer on AI, using Section 5 of the FTC Act to crack down on misleading AI marketing claims, but it operates under existing consumer protection authority rather than any comprehensive AI statute. Without robust federal standards to replace state laws, preemption could create regulatory gaps that leave consumers exposed to algorithmic discrimination, privacy violations, and opaque automated decision-making.
The stakes extend beyond U.S. borders. Companies already navigating the EU's AI Act face a fragmented global compliance landscape. Whether the U.S. settles on a unified federal approach or continues with state-by-state regulation will shape how multinational firms structure their AI governance programs for years to come.
What to Watch Next
Three developments will determine the trajectory of this fight. First, whether congressional committees begin converting the White House framework into actual legislative text -- and whether preemption provisions survive the markup process. Second, the outcome of anticipated legal challenges from Colorado, California, and other states contesting the executive order's constitutionality. Third, the Commerce Department's evaluation of state AI laws, expected to produce formal recommendations on which regulations conflict with federal policy. With Colorado's full enforcement date arriving June 30 and California's CCPA automated decision-making provisions scheduled for January 2027, the clock is ticking on both sides of this regulatory standoff. The only certainty is that AI companies cannot afford to wait for Washington to resolve the question -- compliance with existing state laws remains mandatory until a court or Congress says otherwise.