Trump’s AI Regulation Push Meets State, Congressional Resistance

As the federal government pushes for a single, national approach to artificial intelligence governance, state legislatures are moving in the opposite direction at an unprecedented pace. This clash between federal ambition and state action is defining the early battle over who gets to set the rules for a transformative technology. While the current administration advocates for a minimally burdensome national standard, lawmakers across the country have introduced over 1,200 AI-related bills in a single year, enacting 145 into law. Congress has twice rejected attempts to preempt these state laws, most notably in a 99-1 Senate vote, signaling a deep resistance to centralizing control.

The experience of Utah State Representative Doug Fiefia illustrates the tension. A Republican and former Google employee, Fiefia introduced the Artificial Intelligence Transparency Act earlier this year. His bill, which targeted only the largest frontier AI developers with requirements for safety plans and whistleblower protections, passed a House committee unanimously. Its momentum halted abruptly after the White House sent a letter to state leadership declaring categorical opposition and urging the bill’s defeat without proposing fixes. Fiefia viewed his modest proposal as a matter of principle, arguing for states’ rights even under a Republican administration. The White House treated it as a threat to its broader agenda.

This intervention is part of a structured, three-part federal campaign. It began with Executive Order 14365, signed in December 2025, which established an AI Litigation Task Force within the Department of Justice to challenge state laws in court. The order also directed the Commerce Department to identify burdensome state regulations and linked federal broadband funding to a state’s willingness to avoid what the administration deems onerous AI laws. Commerce delivered its evaluation in March, singling out laws in Colorado, California, and New York for scrutiny. The campaign’s final pillar is a National Policy Framework urging Congress to preempt state laws to avoid a patchwork of fifty different regulations. Former AI czar David Sacks summarized the administration’s view, stating that varied state rules create a difficult compliance landscape for innovators and criticizing blue states for inserting, in his words, “woke ideology” into AI models.

Despite this pressure, state activity has accelerated dramatically. From fewer than 200 bills in 2023, the number skyrocketed to 1,208 in 2025. Major laws in California and Texas took effect at the start of 2026, with Colorado’s algorithmic discrimination ban following later in the year. This flurry reflects a bipartisan state-level consensus that waiting for a gridlocked Congress is not an option. Utah’s Republican Governor Spencer Cox captured this sentiment, advocating for state power to ensure AI benefits humanity without destroying it, and announcing a state-level “pro-human” AI initiative.

For the federal framework to become binding law, however, it needs Congressional approval, which is far from guaranteed. The most comprehensive federal proposal, Senator Marsha Blackburn’s TRUMP AMERICA AI Act, remains a discussion draft. The decisive moment came when the Senate overwhelmingly voted to strip an AI regulatory moratorium from a major funding bill, sending a clear message that federal preemption lacks support. Until courts rule on specific legal challenges, companies must continue navigating the growing web of state regulations.

The financial stakes of this fight are enormous. On one side, the super PAC Leading the Future, backed by Andreessen Horowitz and OpenAI’s Greg Brockman, raised $125 million to support candidates favoring a uniform federal approach. Opposing them, groups like Public First Action, bolstered by a $20 million donation from Anthropic, are pledging millions to back pro-regulation candidates. The tech industry’s total spending to thwart state rules reportedly exceeds $1 billion. A coalition of 36 state attorneys general has also entered the fray, writing to Congress to oppose preemption and argue for the necessity of local protections against scams and deepfakes.

This conflict follows a clear precedent. The administration revoked the previous Biden-era AI executive order, which mandated safety evaluations, within hours of taking office. The trajectory is logical: first, remove federal safety requirements, then actively work to prevent states from creating their own. The outcome is a regulatory vacuum where AI is governed not by deliberate policy but by litigation, executive orders, and corporate influence. This stands in stark contrast to the European Union, where the comprehensive EU AI Act provides a single framework across 27 nations.

Doug Fiefia, whose bill was sidelined, is now running for state senate. He co-chairs a bipartisan AI task force with lawmakers who have direct tech industry experience, representing a new generation that believes understanding AI should inform regulation, not prevent it. Their efforts to fill the governance gap at the state level continue, but the central question remains whether the current vacuum of binding federal rules will become a permanent feature of the American AI landscape.