A federal judge has blocked Montana from enforcing a 2025 law that restricts AI-generated political advertisements, handing critics of state deepfake rules another win in a growing string of courtroom defeats.
U.S. District Judge Susan Watters issued a preliminary injunction on Wednesday, finding that the law's challengers are likely to succeed on their First Amendment claims, according to Reason and multiple Montana news outlets, including Montana Free Press and the Daily Montanan. For more context on this story, see our ongoing AI news.
The Law, and the Mailers That Challenged It
Montana's law made it illegal to sponsor or distribute an election communication within 60 days of voting if the sponsor "knew or should have known" the material is a deepfake, though satire and parody were exempted.
The challenge was brought by Accountability in State Government, a group run by former state lawmaker Dan Bartel. The organization had mailed political attack ads containing doctored images of state politicians — including pictures of officeholders holding pride flags and wearing pronoun badges they never actually posed for.
The mailers prompted complaints from the targeted officials. State Rep. Jennifer Carlson, a Republican from Manhattan, Montana, argued that a mailer showing her with an AI-generated pride flag was created "with the intent to injure the reputation of the candidate or otherwise deceive a voter," according to Reason.
From Dismissed Complaints to an Injunction
The episode took a turn in August, when Montana's Commissioner of Political Practices, Christopher Gallus, dismissed the complaints against the group. Gallus concluded the mailers were sufficiently "humorous or outlandish" to qualify as satire — precisely the exemption the statute contained.
The next morning, Judge Watters heard arguments on Bartel's request to block enforcement of the law altogether. This week she sided with the challengers. The problem, in her view, was the statute's asymmetry: it targeted synthetic media created "with the intent to injure the reputation" of a candidate or party. As Watters put it, the law "treats identically deceptive media differently based solely on whether its message is weaponized to damage or deployed to elevate a candidate" — the kind of viewpoint-based distinction the First Amendment ordinarily does not permit.
Adam Schulman, an attorney with the Hamilton Lincoln Law Institute who is representing Bartel, told Reason that the flaw was fundamental. "You could have AI deepfakes that were fine as long as they were complimentary," Schulman said, offering the hypothetical of a deepfake showing a politician in a Pittsburgh Steelers uniform — flattering, and therefore permitted under the statute. He also argued the law forced regulators into inherently subjective judgments about political imagery, since it relied partly on how a "reasonable person" would understand the material. "People have different artistic sensibilities," Schulman said.
Part of a National Pattern
Montana is hardly alone in discovering the constitutional difficulty of regulating AI-generated political speech. Plenty of states have attempted such restrictions, and the early results have been strikingly consistent.
After conservative YouTuber Christopher Kohls posted an AI-produced parody of then-Vice President Kamala Harris during the 2024 election, California passed Assembly Bill 2839, prohibiting distribution of communications containing "materially deceptive content" produced "with malice." In 2025, a federal court ruled the law unconstitutional and permanently enjoined California from enforcing it.
Hawaii's Act 191, passed in 2024, restricted "materially deceptive" election media. It was challenged by the satirical site the Babylon Bee, along with a Hawaii resident named Dawn O'Kane. In January 2026, U.S. District Judge Shanlyn Park granted summary judgment against the law, finding it "unconstitutionally vague," in part because it "employs ambiguous and subjective standards."
The scoreboard is now hard to ignore: deepfake election laws in California, Hawaii and Montana have all been halted by federal courts within roughly two years of passage.
Why It Matters for AI Policy
The Montana ruling lands as legislators across the country keep reaching for bans on synthetic political media, and as generative tools make fabricated audio and video cheaper and more convincing every year. The court fights are shaping the boundaries of what regulation can lawfully do: intent-based standards that turn on whether a deepfake helps or harms a candidate have fared badly, because they ask the state to punish political messages based on their viewpoint.
Supporters of the laws argue that election integrity demands action on AI forgeries that could deceive voters at scale. The counterargument that is winning in court is procedural and constitutional: the same tools that make a damaging fake also make satire, parody and crude political humor, and laws that cannot reliably tell them apart sweep in protected speech.
What Happens Next
Montana has a month to appeal the injunction, but the state may think twice given the outcomes in California and Hawaii. The broader constitutional question — where exactly the First Amendment lines around AI-generated political media will be drawn — remains unsettled. What the early cases establish is the direction of travel: courts have repeatedly warned that attempts to restrict synthetic political expression run into the same limits that protect parody and unwelcome political speech. For the next round of deepfake bills now being drafted in statehouses, Montana offers the latest evidence that subjective, intent-based standards will face a hard road.
---
Stay Ahead of AIGet the latest AI news, analysis, and breakthroughs — all in one place.
Read more AI news →
