Babylon Bee Sues New Mexico over Satire Disclaimer

Warning labels for AI-generated content make sense. But should they also apply when artificial intelligence is used to create satire or parody? The question is whether a joke can survive a disclaimer.

Question about labeling AI-created jokes.

Do you have to label a joke or a piece of satire just because it was created using AI? Photo: Statement / AI

US satire website The Babylon Bee filed a lawsuit against members of the New Mexico State Ethics Commission on 11 August. The company is seeking to prevent a law passed in 2024 from being applied to its AI-generated political satire. Taken literally, the law could require a joke to be labeled as a joke simply because AI was used to create it.

New Mexico amended its Campaign Reporting Act through House Bill 182 (HB 182). Political advertising containing “materially deceptive media” must now carry a notice stating that an image, video or accompanying audio has been manipulated or generated using artificial intelligence.

The provision covers media showing a person, without their consent, doing or saying something that never happened, where the material was created wholly or partly using artificial intelligence. In videos, the notice must remain visible throughout.

Warning: Satire

The law does not prohibit AI-generated satire. The dispute instead concerns whether the state may compel satirists to attach a prescribed disclaimer to their work. The Babylon Bee invokes the First Amendment to the US Constitution and its protection against government-compelled speech, commonly known as “compelled speech”.

AI Can Already Control Much of Your Smartphone

You might be interested AI Can Already Control Much of Your Smartphone

The irony is that the disclosure requirement expressly applies to satire. The law specifically mentions satire and parody. A satirist could therefore be required to point out that an obviously fictional and deliberately comic depiction is, in fact, fictional.

Severe Penalties

Stricter rules also apply during the 90 days before an election. Anyone who knowingly distributes manipulated material in order to influence voters through a false depiction of a person can face legal consequences. The material must be reasonably likely to affect how citizens vote. Here, too, an AI disclaimer provides an exemption from the rule.

The law also provides for substantial penalties. Violations of the disclosure requirement can result in civil fines of up to $1,000 per violation and up to $20,000 in total. A knowing and willful violation of the special 90-day rule is a misdemeanor on first conviction and a fourth-degree felony on a second conviction.

The Babylon Bee’s current lawsuit was prompted by a satirical travel video about New Mexico published on 4 August. According to the complaint, the video also features David K. Thomson, a justice of the New Mexico Supreme Court.

A real image of Thomson is followed by AI-generated scenes in which a fictional version of the justice announces rulings against various constitutional rights. Thomson serves on the state’s highest court. His term expires at the end of 2026.

A Pre-Emptive Challenge

The Babylon Bee fears that the video could fall under the law. The website has not been penalized over its publication. Through its pre-enforcement lawsuit, its operators are seeking to prevent the provisions from being applied to the video.

https://www.youtube.com/watch?v=lttDXMHQ3n4

They argue that the state is interfering with political speech while also forcing the creator to include a government-prescribed message in its own work.

The Babylon Bee’s argument has support from an unlikely source: New Mexico Attorney General Raul Torrez. At the request of Governor Michelle Lujan Grisham, he reviewed the law in 2025.

Torrez concluded that a general disclosure requirement for AI-generated content was constitutional in principle. In his view, a brief factual notice could serve the legitimate purpose of protecting voters from deceptive AI-generated material.

A Constitutional Problem

Torrez reached the opposite conclusion when it came to satire and parody. A depiction that could reasonably be understood as satire or parody, he argued, is by definition unlikely to mislead a reasonable viewer about actual events.

In such cases, he found an insufficient connection between the state’s interest in protecting voters and the disclosure requirement. According to his constitutional analysis, that portion of the law was likely unconstitutional on its face.

The State Ethics Commission rejected the allegations after the lawsuit was filed. It told Bloomberg Law that the provisions had never been enforced against political satire or parody.

https://twitter.com/ReclaimTheNetHQ/status/2089073550657560709

The Babylon Bee counters that it should not have to risk punishment first. The credible threat of enforcement, it argues, may already deter protected political speech.

A Disclaimer for a Joke

The New Mexico dispute is not the first of its kind in the US. Federal courts have previously struck down comparable rules in California and Hawaii as unconstitutional. In Hawaii, the state allowed the deadline for an appeal to expire, while the California ruling has been appealed.

There is little dispute that the state has a legitimate interest in protecting voters from deceptive AI manipulation. The question is how far it may go in restricting political speech in pursuit of that goal.

The courts must now decide whether the same disclaimer may be required where the obvious unreality of the depiction is precisely what makes it satire. May the state require the clown to hold up a warning sign before telling his own joke simply because AI helped create it?