BILLINGS, Mont. — A federal judge has temporarily barred Montana officials from enforcing the state’s election-deepfake law against one political action committee and its leader, finding that they are likely to succeed on a First Amendment challenge.

U.S. District Judge Susan Watters issued the preliminary injunction September 16 in Accountability in State Government v. Knudsen. The order protects Accountability in State Government and former Montana lawmaker Dan Bartel from investigation or prosecution under the law while their case continues.

The ruling is narrow. It does not suspend the law statewide, strike it from Montana’s statutes or finally decide that the law is unconstitutional. The preliminary injunction applies to the named plaintiffs, and Montana may appeal.

Why the judge intervened

Montana defines an election “deepfake” as artificial-intelligence-generated or synthetic content depicting a candidate or political party with an intent to damage that candidate’s or party’s reputation or otherwise deceive voters.

The judge concluded that the challengers are likely to show that this definition discriminates by viewpoint: deceptive AI media intended to harm a candidate can be restricted while materially similar synthetic media intended to benefit a candidate may fall outside the definition.

The court’s finding concerns the challengers’ likelihood of success at this preliminary stage. It is not a final merits judgment, and the court did not approve the truthfulness or substance of the political materials involved in the dispute.

What Montana’s law requires

Within 60 days of voting beginning, covered campaigns, political committees and other entities generally may not knowingly sponsor the creation or distribution of a deepfake involving a candidate or political party unless the communication carries the statute’s disclosure.

For audio, the warning must be clearly audible and intelligible, last at least eight seconds and appear at the beginning or end of the communication. The prescribed language identifies the material as significantly edited by AI and says it falsely appears authentic or truthful.

What the ruling means for radio

Montana’s statute contains specific protections for broadcasters. A radio or television station is exempt from the law’s prohibition and penalties when it is paid to carry a deepfake advertisement or when federal law requires it to air an advertisement from a legally qualified candidate.

That broadcaster exemption does not necessarily protect the campaign, PAC, agency or other organization that created or purchased the advertisement. Stations should continue requiring political advertisers to identify AI-manipulated material, document who supplied each spot and provide the applicable disclosure.

The law also permits bona fide newscasts, interviews, documentaries and on-the-spot news coverage to use deepfake material when the broadcast clearly acknowledges—through context or a disclaimer—that the content may be AI-generated and may not accurately portray the depicted person’s speech or conduct.

The larger media-law question

The decision adds to the constitutional scrutiny facing state attempts to regulate synthetic election media. The immediate question is not whether states may address deceptive AI content at all, but whether a law may regulate it differently depending on whether the message helps or harms a candidate.

For radio companies operating in multiple states, the practical challenge remains a patchwork of different definitions, disclosure language, timing rules, exemptions and enforcement procedures. The Montana order changes enforcement only for the two plaintiffs; it does not create a nationwide rule for broadcasters or political advertisers.

Primary sources

This report provides general information and is not legal advice.