The AI Regulation Showdown: Federal vs. State Power

"We need one central source of approval" — President Trump, on the need for a unified national AI strategy.

A complex debate over the regulation of artificial intelligence (AI) is unfolding in the United States, characterized by a conflict between federal and state authorities and significant political spending by AI industry figures. The Trump administration has issued an executive order aimed at limiting state-level AI regulations, while individual states continue to pursue their own laws, and AI-linked spending in congressional races has reached new highs.

Federal Executive Order Challenges State AI Laws

President Donald Trump signed an executive order outlining a strategy to preempt state and local AI regulations. The order includes several key directives:

• AI Litigation Task Force: The Department of Justice is directed to establish a task force to initiate lawsuits against states regarding their AI legislation.
• Agency Collaboration: The Federal Trade Commission (FTC) and Federal Communications Commission (FCC) are instructed to work with the Justice Department to address state regulations deemed "onerous."
• Federal Funding Review: The Commerce Secretary is directed to investigate the authority to withhold federal rural broadband funding from states that have enacted AI laws considered unfavorable.

Administration's Rationale

The administration argues that competitive pressure from China requires a unified national approach, not a patchwork of state rules.

White House AI adviser David Sacks stated the order provides tools to address state regulations characterized as "onerous," but clarified it would not oppose regulations concerning children's safety. The administration has released a framework for potential congressional legislation that would preempt state laws considered "too burdensome" while allowing some rules.

Industry Perspective

Major AI companies — including OpenAI, Google, Meta, and Anthropic — have expressed a preference for national legislation, arguing that a patchwork of state-level rules could hinder innovation and affect the United States' competitive standing.

State-Level AI Regulation Efforts Continue

Despite the executive order, more than 1,000 state legislative proposals addressing AI have been introduced. Key state actions include:

• California: Governor Gavin Newsom signed a bill requiring major AI developers to submit plans for mitigating risks, suggesting this law could serve as a model for federal legislation.
• New York: Passed legislation requiring major AI developers to report dangerous incidents to the state.
• Colorado and Florida: Have passed or are advancing laws regulating AI development, including measures related to child safety and transparency.
• Utah: A Republican state representative's bill requiring child safety protocols failed to proceed after White House opposition.

State Lawmakers' Perspective

"Congress faces gridlock, giving states an opportunity to act" — Republican State Rep. Doug Fiefia of Utah.

State lawmakers argue that congressional gridlock necessitates state action. Republican State Sen. Angela Paxton of Texas noted that the lack of federal legislation requires states' ability to pass laws. Republican State Sen. Tracy Pennycuick of Pennsylvania highlighted states' ability to respond quickly to problems.

Political Reactions and Divisions

The executive order has drawn criticism from some typically pro-Trump organizations involved in child safety. Michael Toscano of the Institute for Family Studies called it a "lost opportunity." Adam Billen of Encode expressed concern about a potential "chilling effect" on states' willingness to protect residents.

California Governor Gavin Newsom criticized the order as an attempt to preempt state laws protecting residents from unregulated AI technology. Julie Scelfo of Mothers Against Media Addiction argued that preventing states from enacting AI safeguards undermines their rights.

The issue has created divisions within the Republican party itself.

Senator Josh Hawley (R-MO) opposed an attempt to insert AI preemption into the defense spending bill. Governor Spencer Cox (R-UT) expressed preference for an order that would not bar state laws. Governor Ron DeSantis (R-FL) stated an executive order "doesn't/can't preempt state legislative action."

Record AI Industry Spending in Congressional Races

AI-linked super PACs have spent $43.3 million on US congressional races in the 2025-2026 election cycle — according to OpenSecrets.

A significant portion of this spending occurred in New York's 12th Congressional District Democratic primary.

New York Primary Outcome

State Assemblyman Micah Lasher won the Democratic primary, receiving 39% of the vote. Lasher was supported by Governor Kathy Hochul, former Mayor Michael Bloomberg, and retiring Representative Jerry Nadler. He is expected to win the general election in the deep-blue district.

State Assemblyman Alex Bores, who focused his campaign on AI regulation and co-sponsored New York's RAISE Act, received 35% of the vote. Jack Schlossberg, grandson of former President John F. Kennedy, finished third with 10.8%.

Spending Breakdown

The primary attracted unprecedented outside spending from competing AI factions:

• Anti-Bores Spending: Think Big PAC, affiliated with Leading the Future (backed by Andreessen Horowitz, OpenAI co-founder Greg Brockman, Perplexity, and Palantir's Joe Lonsdale), spent at least $8 million opposing Bores. Leading the Future and its affiliated super PACs spent $23.5 million on races nationwide.
• Pro-Bores Spending: Public First Action (which received $20 million from Anthropic) and allied super PACs spent $11 million supporting Bores. Crypto billionaire Chris Larsen also spent over $3 million supporting Bores.
• Lasher Support: Former Mayor Michael Bloomberg spent at least $10 million supporting Lasher, a former employee.

Industry Factions

The spending reflects deep divisions within the AI industry. OpenAI and its investors generally oppose strict state-level regulation, favoring federal preemption. Anthropic supports stronger safeguards. Leading the Future co-leader Josh Vlasto stated the PAC supports a national regulatory framework that creates jobs, helps compete with China, and includes safety guardrails.

Public Opinion and Future Outlook

Approximately 80% of US residents expressed concern about AI, with about 75% saying the government is not doing enough to regulate it. Approximately 90% of Democrats and 60% of Republicans polled wanted more government involvement.

The executive order is expected to face legal challenges. Tech policy researchers suggest the administration may lack authority to restrict state regulation without specific congressional legislation. John Bergmayer of Public Knowledge argued that the administration is attempting to bypass Congress, citing a 2023 Supreme Court decision that upheld a state's right to regulate its industries.

On Capitol Hill, Republican allies of the president are working on legislation, such as the TRUMP AMERICA AI Act, to codify the administration's agenda. However, prior attempts to introduce AI preemption into federal legislation have failed.