Trump Administration Pushes to Block State AI Regulations as States Accelerate Legislation

At a glance:

• The Trump administration is using a DOJ task force, Commerce evaluations, and legislative pressure to prevent states from regulating AI.
• 1,208 AI bills were introduced in 2025 across all 50 states, with 145 enacted into law.
• Congress has rejected federal preemption of state AI laws twice, including a 99-1 Senate vote to remove a moratorium.

The Federal Architecture

The Trump administration’s campaign against state AI regulation is built on three pillars: an executive order, a Commerce Department evaluation, and a proposed national framework. Signed on 11 December 2025, Executive Order 14365 established an AI Litigation Task Force within the Department of Justice, set to challenge state laws in federal court starting January 2026. The task force will target laws deemed "burdensome" under the Commerce Clause or conflicting with federal preemption. The order also directed the Commerce Department to publish an evaluation of state AI laws by 11 March 2026, identifying which regulations might hinder interstate commerce. This evaluation, released on 11 March, flagged laws in Colorado, California, and New York for scrutiny. The FTC was tasked with issuing guidance on when state laws are preempted by the FTC Act. Additionally, federal broadband funding was conditioned on states avoiding "onerous" AI laws, though child safety, data center zoning, and state procurement were exempt from preemption.

The Commerce Department’s evaluation, published exactly on the March deadline, listed specific state laws it deemed problematic. For example, Colorado’s algorithmic discrimination rules and California’s transparency requirements were highlighted as potential barriers to national consistency. The evaluation feeds directly into the DOJ task force, which is expected to file lawsuits by summer 2026. Legal battles could take two to three years to resolve, creating uncertainty for both states and tech companies. The administration’s National Policy Framework for AI, released on 20 March, further emphasized preemption, urging Congress to establish a "minimally burdensome national standard" to replace what it called "fifty discordant state regulations."

States Are Accelerating Legislation

While the federal government resists state action, states have embraced AI regulation at an unprecedented pace. In 2023, fewer than 200 AI bills were introduced across state legislatures. By 2024, that number surged to 635 across 45 states, with 99 enacted. In 2025, all 50 states introduced at least one AI-related bill, totaling 1,208 proposals, with 145 becoming law. This includes California’s Transparency in Frontier Artificial Intelligence Act (effective 1 January 2026), which mandates safety and child-protection plans from frontier AI companies. Texas passed its Responsible Artificial Intelligence Governance Act on the same date. Colorado’s AI Act, which bans algorithmic discrimination, was delayed until 30 June 2026. The surge reflects bipartisan consensus at the state level, with lawmakers arguing that federal inaction necessitates local solutions.

Utah’s experience illustrates the tension. Doug Fiefia, a Republican state representative, introduced House Bill 286—the Artificial Intelligence Transparency Act—in early 2025. The bill required frontier AI developers (those using at least 10^26 floating-point operations) to publish safety and child-protection plans, with a $1 million penalty cap. It passed a House committee unanimously but was killed by the White House. On 12 February, the White House Office of Intergovernmental Affairs sent a letter to Utah’s Senate Majority Leader, calling the bill "unfixable" and opposing it "categorically." Fiefia, a former Google salesperson, defended his stance, stating that states must act to protect citizens from risks like sexualized chatbots targeting minors. His bill, though modest by AI legislative standards, was framed by the administration as existential. This dynamic has played out across states, with legislators pushing back against federal preemption.

Congressional Inaction and the Preemption Debate

Despite the administration’s push, Congress has repeatedly rejected efforts to preempt state AI laws. The most comprehensive federal proposal, Senator Marsha Blackburn’s TRUMP AMERICA AI Act, was released as a 291-page discussion draft on 18 March. It would impose a duty of care for high-risk AI systems, require data use disclosures, repeal Section 230 of the Communications Decency Act, and create a liability framework for AI developers. Crucially, it would preempt state laws on frontier AI catastrophic risks but largely allow state digital replica laws. However, the bill remains a draft and has not been formally introduced. The One Big Beautiful Bill Act, which originally included a five-year moratorium on state AI regulation tied to federal broadband funding, was stripped of its preemption clause after a 99-1 Senate vote. Only Senator Thom Tillis of North Carolina voted to retain the provision. The final law, signed on 4 July, left state regulation intact.

The administration’s framework requires Congressional action to gain legal force, as the executive order itself does not preempt state laws. Until courts rule on specific challenges, companies must comply with state regulations. This has created a legal gray area, with states like Colorado’s attorney general pledging to challenge the executive order in court. The contrast with the EU’s AI Act, which enforces a single regulatory framework across 27 member states, underscores the US approach: no binding federal standard and an active campaign to block state action.

Financial and Political Leverage

The lobbying battle over AI regulation has attracted massive funding. Leading the Future, a super PAC launched by Andreessen Horowitz and OpenAI’s Greg Brockman, raised $125 million in 2025 and had $70 million remaining at year-end. It supports candidates favoring federal preemption and uniform regulation. On the other side, Anthropic donated $20 million in February 2026 to Public First Action, a bipartisan group backing candidates for AI safeguards. Public First’s network has pledged $50 million for pro-regulation candidates. The tech industry reportedly spent over $1 billion to prevent state regulation, reflecting the stakes for companies facing varying state rules.

A bipartisan coalition of 36 state attorneys general opposed preemption, arguing that state laws are essential to address risks like deepfakes, scams, and harmful AI interactions, particularly for vulnerable groups. Colorado’s attorney general has committed to litigating the executive order, joining a growing legal front against federal preemption efforts.

The Precedent of Federal Withdrawal

The administration’s approach is a direct reversal of Biden-era policies. Within hours of taking office on 20 January 2025, the Trump administration revoked Executive Order 14110, which required AI safety evaluations and data sharing with the government. The replacement order, titled "Removing Barriers to American Leadership in Artificial Intelligence," emphasized deregulation. This shift from federal safety mandates to blocking state regulation has a logical progression: if the federal government will not act, and states are barred from acting, AI regulation may vanish entirely.

Doug Fiefia, now running for state senate, embodies the state-level resistance. His opponent, the incumbent who helped kill his bill, claimed it would "drive Utah out of the AI innovation business." Fiefia co-chairs the AI task force of the Future Caucus, a bipartisan group of tech-savvy lawmakers. The question remains whether the regulatory vacuum created by federal inaction will persist long enough to become permanent.