Regulating Political Deepfakes: First Amendment Lessons from Hawaii’s Unconstitutional Effort

by Clay Calvert, AEI Nonresident Senior Fellow, Technology Policy Studies, March 18, 2026

Federal Judge Shanlyn Park recently delivered an important First Amendment victory for satiric commentary in Babylon Bee v. Lopez, pushing back against an overly broad government response to political deepfakes. Park permanently enjoined a Hawaii law criminalizing the reckless distribution of “materially deceptive media” that “risk[s] . . . harming the reputation or electoral prospects of a candidate in an election or changing the voting behavior of voters in an election.”

Act 191 targets content generated by artificial intelligence and digital technology that depicts candidates “engaging in speech or conduct in which . . . [they] did not in fact engage” and that “would cause a reasonable [person] to believe . . . [they] engaged in the speech or conduct depicted.” Additionally, the verboten content must advocate for or against depicted candidates, akin to an “advertisement.”

Act 191 exempts manipulated content that features a “clearly visible” disclaimer explaining it is fake. Democratic Governor Josh Green signed the measure into law in 2024.

Park agreed that Hawaii has “a compelling interest in regulating political deepfakes for the purpose of protecting the State’s electoral integrity—an essential democratic function.” The problem is that Act 191 isn’t narrowly tailored—isn’t sufficiently confined in scope—to serve that interest. It therefore fails the strict scrutiny test for examining the constitutionality of content-based statutes (ones that regulate speech about some subjects but not others).

Although arguably well intended to shield voters from what Act 191 calls “disinformation and misinformation,” the law fails to adequately account for the competing free-speech interests of satirists who skewer politicians. Here are two constitutional principles lawmakers should remember when drafting deepfake statutes.

Narrow Tailoring. If there’s an alternative way to effectively serve a compelling government interest (here, electoral integrity) that restricts less speech—impinges less on others’ expressive interests—than the one(s) lawmakers selected, then a content-based law isn’t narrowly tailored and violates the First Amendment. Park concluded that “counter speech appears to be a viable, less restrictive alternative.”

Counter speech—a venerable doctrine—was imported into First Amendment law nearly a century ago in Whitney v. California. It holds that “if there be time to expose through discussion the falsehood and fallacies, to avert the evil by the processes of education, the remedy to be applied is more speech, not enforced silence.”

Counter speech remedies for political deepfakes that Park deemed viable include “crowd-sourced fact checking,” plus Hawaii “counter[ing] deceptive speech with factual speech of its own,” and creating a database tracking materially deceptive content. Furthermore, Park reasoned that Hawaii could launch an educational campaign to increase “the digital and political literacy of the electorate” in spotting deceptive political content. In sum, counter speech and educational initiatives provide less speech-restrictive ways of protecting voters from deepfake deception that don’t criminalize satire.

Safeguarded Satire. The First Amendment presumptively protects political satire about public figures. In 1988, the US Supreme Court affirmed pornographer Larry Flynt and Hustler magazine’s First Amendment right to satirize the Reverend Jerry Falwell as a “drunk and immoral . . . hypocrite who preaches only when . . . drunk” and who lost his virginity “during a drunken incestuous rendezvous with his mother in an outhouse.” Hustler’s satiric vehicle was a parody of actual Campari liqueur advertisements. Falwell, then a “high-profile figure on the political scene” who founded Liberty University and the Moral Majority, was a prime satirical target for a pornographer like Flynt.

In declaring Falwell a public figure and ruling against his claim for intentional infliction of emotional distress, the Supreme Court began from the premise that “at the heart of the First Amendment is the recognition of the fundamental importance of the free flow of ideas and opinions on matters of public interest and concern.” It observed that “despite their sometimes caustic nature, from the early cartoon portraying George Washington as an ass down to the present day, graphic depictions and satirical cartoons have played a prominent role in public and political debate.” The Court concluded that “from the viewpoint of history, it is clear that our political discourse would have been considerably poorer without them.”

This is exceedingly relevant for the Lopez plaintiffs. The Babylon Bee wryly dubs itself “the world’s best satire site, totally inerrant in all its truth claims.” Plaintiff Dawn O’Brien is a Honolulu resident who “regularly posts about politics, elections, and religion on her public Instagram and Facebook accounts.”

Both plaintiffs deliberately create and distribute fake content about politicians to satirize them and their policies. O’Brien plans this year to distribute “an AI-generated image of President Donald Trump endorsing Governor Green for reelection and AI-generated images of Governor Green holding signs that read ‘Free speech is cancelled’ and ‘Satire and parody require labels.’”

Thanks to Judge Park’s astute ruling blocking Act 191, O’Brien may freely do so.