Here Are the States Making Big Moves Toward AI Regulation in 2025

Few political observers are counting on Washington to implement new AI laws this year, but several states are making moves toward regulating the fast-growing tech. 

If those proposals succeed, they’ll add to the rapidly multiplying number of state laws aimed at cracking down on what some lawmakers see as an emerging threat to free and fair elections. But the new bills also create hurdles for political professionals, who already have to contend with a patchwork of state laws governing the use of AI in elections.

Here’s a look at a few of the states making big moves toward AI regulation in 2025:

Texas

A 44-page bill introduced by Republican state Rep. Giovanni Capriglione is the most sweeping AI proposal on the table this year. If it passes, it would enact some of the most comprehensive regulations on AI technologies to date, ranging from imposing new transparency requirements to protecting consumers from algorithmic discrimination.

What has some political professionals worried is the sheer breadth of the proposal, dubbed the Texas Responsible AI Governance Act, or TRAIGA. 

While the bill covers only “high-risk” AI systems – those that play substantial roles in consequential decision making, it would also create broad AI disclosure requirements and ban “subliminal techniques” or “purposefully manipulative or deceptive techniques.” How that affects the political world remains an open question.

Arkansas

State lawmakers on both sides of the aisle have already filed bills this year that would crack down on deceptive or fraudulent AI-generated content in elections. 

One proposal, introduced by Republican state Rep. Scott Richardson, would make it a misdemeanor to deploy “deceptive and injurious” deepfakes during the course of an election, while a Democratic-led bill would establish civil fines for the use of such AI-generated content. 

The bills show a concerted, bipartisan effort to clamp down on the growing use of AI systems to spread misinformation in elections – a big move for Arkansas, which was one of the few states where legislators didn’t consider any AI-related proposals in 2024.

Nevada

A new bill introduced in the Silver State last week would require companies that offer AI services to register with the state’s consumer protection office. As part of that, those companies would have to provide information on how and where the data is stored, while search engines would have to allow users to opt into AI functions.

Companies would also be required to develop a framework for addressing misinformation, fraud, hate speech and bias in their AI systems. 

Another proposal in Nevada introduced last month would create disclosure requirements for political campaigns that use “synthetic media” to alter or create an image. Violators could be fined as much as $50,000 under the proposal.

North Dakota

North Dakota lawmakers are also moving to create disclosure requirements for the use of AI in political communications and advertising. 

The proposal, introduced by Republican state Rep. Jonathan Warrey, would require a “prominent disclaimer” on any AI-generated content used for political purposes. It would also make any violations a Class A misdemeanor. 

A separate bill that would outlaw the use of fraudulent deepfake videos or images failed in the North Dakota state House last month.

Virginia

A proposal that would require all AI-generated electioneering communications to feature a disclaimer is making its way through the Virginia General Assembly. Both the state House and Senate approved versions of the bill earlier this month. 

Under the proposal, any “synthetic media” used in political communications would have to feature a disclosure statement noting that the “message contains synthetic media that has been altered from its original source or artificially generated and may present conduct or speech that did not occur.”

Violators could face a financial penalty or a misdemeanor charge.