A federal judge denied Elon Musk's xAI a temporary restraining order against Minnesota's sweeping nudification ban, letting a law with $500,000 civil penalties take effect Aug. 1 — and shrinking the carve-outs for legitimate expression in the process.
The real problem is real: creeps using AI to strip real people's clothes off photos is exploitation, and it deserves a legislative response. But Minnesota didn't just target the creeps. The law bans accessing, down, or using any website or app capable of nudification — and it bans advertising or promoting such products, too. That's a net wide enough to catch a lot more than bad actors.
U.S. District Judge Donovan Frank ruled that xAI's delay in filing — nearly three months after the law was signed, and just three days before its effective date — "suggests that harm is not immediate." A hearing on a preliminary injunction is set for Aug. 19. In the meantime, the restrictions signed by Gov. Tim Walz stay on the books.
xAI argued in court documents that the law "imposes an overbroad, content-based ban on free speech and the tools of visual expression in a clumsy attempt to prohibit 'nudification.'" The company stressed it "does not contest Minnesota's interest in prohibiting the dissemination of artificially generated nude images of real people without their consent" — but said the statute "extends far beyond that goal, exposing a wide array of protected speech to civil liability and government sanctions."
xAI also noted it already prohibits users from generating non-consensual sexualized images and has sued users who evade its technological blockers, according to FOX 9.
Walz, for his part, responded with the kind of gravitas Americans have come to expect from their governors: "See you in court, creep," he posted on X, referring to Musk, as UPI reported.
UPI framed the story around scrutiny xAI faced after a December release of Grok generated thousands of sexualized images, including some that appeared to be of minors. FOX 9 noted the company's existing safeguards and the judge's direction for parties to meet and confer on a potential agreement. Neither outlet pressed hard on the constitutional question at the center: whether a law written this broadly can survive strict scrutiny.
Minnesota lawmakers held a press conference July 29 expressing confidence the ban would survive the challenge. When both parties and the press agree a law is just common sense, that's usually when the public should read the fine print.
The law's penalties — at least $500,000 per unlawful access — aren't targeted at individual bad actors. They're aimed at platforms and tools. That's the tension: you can oppose deepfake exploitation and still recognize that a content-based ban on software tools, enforced by ruinous civil penalties, is a blunt instrument that will chill legitimate speech before it stops a single predator.
The Aug. 19 hearing will tell us whether the courts are willing to do what the legislature wouldn't — narrow the law to hit the problem without gutting the First Amendment.