← Back to all posts
News

xAI Sued Minnesota Over Its Nudify Ban. Walz: 'See You in Court, Creep'

July 29, 2026 · 03:12 UTC · News
xAI Sued Minnesota Over Its Nudify Ban. Walz: 'See You in Court, Creep'

TL;DR

On Monday, July 27, xAI sued Minnesota Attorney General Keith Ellison in federal court to stop HF 1606, the first state law in the country to ban nudification apps, from taking effect on August 1. The complaint calls the statute "an overbroad, content-based ban on free speech and the tools of visual expression in a clumsy attempt to prohibit 'nudification.'" The law carries fines of $500,000 per violation, passed the Minnesota House 132-1 and the Senate 65-0, and Governor Tim Walz responded to the lawsuit on X with "See you in court, creep." Behind the theatrics sits the question every image-model operator should care about: can a state attach strict per-output liability to a generative model, no matter what safeguards the operator runs?


The law xAI wants dead

Minnesota's legislature was the first in the country to ban nudification technology outright: services that alter or generate an image or video to depict an intimate part of an identifiable person that was not in the original. The law targets the companies that operate and host these tools rather than the individuals who use them, and it lets both victims and the state attorney general sue.

The price tag is what makes it bite: $500,000 per violation. It passed with the kind of margin state legislatures usually reserve for naming bridges, 132-1 in the House and 65-0 in the Senate, and Walz signed it earlier this year. It goes live Saturday, August 1.

hf 1606 floor votes, 2026 session House yes132 House no1 Senate yes65 Senate no0
197 votes for, one against, across both chambers. The lone no vote came in the House.

The origin story matters for how this plays in court. Per KSTP, state Sen. Erin Maye Quade championed the bill after learning about a man who used social media photos to create sexualized images and videos of more than 80 women he knew, without their consent. Rep. Jessica Hanson, the chief House sponsor, answered the lawsuit this week with "Nudification is simply an indefensible tool of AI."

xAI's case: everyone's a nudifier now

The complaint, filed in the District of Minnesota, does not defend nudification. xAI says it takes no issue with outlawing the dissemination of artificially generated nude images of real people without their consent. Its argument is that HF 1606 sweeps in far more than that.

Exhibit-grade specifics, per CBS Minnesota and Gizmodo: the statute's definition of "intimate part" covers male and female breasts and inner thighs, which xAI reads as banning generated images of shirtless men, people in swimsuits, and shorts. The complaint says the law reaches artistic, scientific, political, satirical, educational, medical, and religious imagery, and it cites an AI-generated picture President Trump himself posted, depicting the president, the vice president, and cabinet members in bathing suits in the Reflecting Pool. A shirtless cabinet in the Lincoln Memorial Reflecting Pool is now a load-bearing exhibit in American constitutional law, which feels about right for 2026.

The structural objection is the one that matters beyond Minnesota: liability is strict. The fine lands regardless of the operator's safeguards, the user's intent, or even the depicted person's consent. It is speed-camera law for models: the ticket arrives no matter who was driving or how good your brakes are, and each ticket is $500,000.

how hf 1606 assigns liability: strict, per output user promptsany nudify tool nudified image ofidentifiable person operator owes$500,000 no intent test :: safeguards don't offset :: victims and the AG can sue
Liability attaches to the output itself. By the complaint's own arithmetic, 100,000 violating images would be $50 billion in exposure.

The awkward resume

xAI is a strange champion for this fight, and the complaint knows it. Per the Minnesota Reformer, xAI tells the court it strictly prohibits generating nude or sexualized images of people without consent, has suspended more than 50,000 accounts, and has filed more than 70,000 reports to the National Center for Missing and Exploited Children, which it says contributed to 244 arrests. It has even sued its own users for evading its filters, a lawsuit we covered earlier this month.

The other side of that ledger is why Minnesota legislators moved in the first place: Grok's image tools spent months in headlines for generating sexualized imagery of real people at scale before xAI tightened its blockers, including, per Gizmodo's reporting, imagery of minors. When your moderation statistics double as the prosecution's opening slide, citing them is a bold strategy.

Why this is bigger than Grok

Strip out Musk and this is the first serious federal court test of a state statute that regulates what a generative model may output, enforced against the model's operator. Every AI content law that follows will be drafted in the shadow of whatever the District of Minnesota decides.

If HF 1606 survives, "we moderate" stops being a legal defense in Minnesota, and per-output strict liability becomes a template other statehouses can copy the way they copied election-deepfake laws. The practical compliance answer for image-model operators becomes geo-fencing states, not shipping better filters, because better filters do not reduce the fine. If it falls, states lose their sharpest enforcement tool against nudify apps and get sent back to intent-based statutes that are far harder to prosecute.

The timing is not subtle either. CBS notes the Trump administration has already said it wants to challenge state-level AI laws over patchwork concerns, and the EU's AI Act transparency fines begin this same weekend. The next few days are a compressed preview of the decade's central AI policy fight: who gets to regulate model outputs, and at what layer of the stack.

What happens before Saturday

xAI is asking the court to declare the law unconstitutional and block enforcement, which means the immediate question is whether an injunction lands before August 1. Ellison is not conceding an inch: "AI nudification robs the target of their dignity and can cause immense harm on an emotional, personal, and professional level," he said, adding there are no worthy debates to have on this issue.

And Walz, who signed the bill, compressed the state's litigation posture into five words on X: "See you in court, creep." Some constitutional scholars quoted in early coverage think xAI's overbreadth arguments have real teeth, which is the uncomfortable part for everyone rooting against Grok here. A 197-1 law can still be a sloppily drafted one, and sloppy drafting is exactly what First Amendment litigators eat.

Key Takeaways

  • xAI filed a First Amendment challenge on July 27 in the District of Minnesota against AG Keith Ellison to stop HF 1606, the nation's first nudification-app ban, before it takes effect August 1.
  • The law fines operators $500,000 per violating image, with strict liability that ignores safeguards and intent, and lets both victims and the attorney general sue.
  • xAI argues the definition of "intimate part" (breasts, inner thighs) sweeps in swimsuits, shirtless depictions, satire, and political imagery, citing a Trump-posted AI image as an exhibit.
  • The complaint touts xAI's own enforcement record: 50,000+ suspended accounts and 70,000+ NCMEC reports it says contributed to 244 arrests.
  • Minnesota passed the law 132-1 and 65-0 after a case involving sexualized fakes of more than 80 women; Walz's response to the suit was "See you in court, creep."
  • This is the first real court test of per-output strict liability for a generative model; the ruling will shape every state AI content law that follows.

Sources: CBS Minnesota, KSTP, Minnesota Reformer, KTTC, Gizmodo

AIxAIGrokRegulationLegalImage GenerationDeepfakesPolicy
CONSOLE
$