Minnesota’s deepfake ban just took effect and xAI couldn’t stop it

xAI waited almost three months to sue over a law it called unconstitutional, then asked a federal judge to block it three days before it kicked in. That timing is what sank the request.

US District Judge Donovan Frank refused to stop Minnesota from enforcing its new law banning apps and websites capable of generating nonconsensual intimate imagery, according to NBC News. The law took effect Aug. 1.

The judge did the math on the calendar

Frank wasn’t convinced xAI faced immediate harm. He noted the company filed its lawsuit July 29, 2026, nearly three months after the law was signed and just three days before enforcement began.

“Such a delay in bringing the action and the motion suggests that harm is not immediate,” the judge wrote.

That’s a procedural loss, not a ruling on the merits. The case continues, and the court will hold a hearing Aug. 19 on whether to grant xAI a preliminary injunction against the law, the first of its kind in the country.

$500,000 per image, and the fine lands on the developer

The penalty structure is what makes this law different from the usual state-level attempt at regulating synthetic media. Minnesota doesn’t just ban the tools. It fines developers $500,000 every time a user generates a nonconsensual intimate image using their product.

The definition is narrower than the headline suggests. A “nudified” image or video is one “altered or generated to depict an intimate part not depicted in an original unaltered image or video of an identifiable individual.”

And it only applies when the result is “so realistic that a reasonable person would believe that the intimate part belongs to the identifiable individual.” Cartoons and obvious fakes aren’t the target. Convincing ones are.

xAI says it's not defending nudification

The company’s complaint calls the law an “overbroad, content-based ban on free speech and the tools of visual expression in a clumsy attempt to prohibit ‘nudification.'”

xAI denied it was contesting Minnesota’s effort to stop the spread of artificially generated nude images of real people. It argued instead that the law “extends far beyond that goal, exposing a wide array of protected speech to civil liability and government sanction.”

It’s a familiar argument: the state aimed at a real harm and hit a wide range of lawful expression on the way. Whether it holds up gets tested Aug. 19.

Grok’s track record is the uncomfortable part

xAI is arguing this in a First Amendment frame, but the factual backdrop is specific to its own product. Grok drew attention earlier this year for complying with requests to turn photos of real women and children into sexualized images.

The company made policy changes by mid-January to address it. NBC News reports Grok was still undressing real people as recently as April.

Three months of fixes, and the behavior persisted. That gap between the stated policy and what the model actually did is the thing Minnesota’s law is designed to price, at $500,000 a shot, and it’s the detail xAI’s lawyers will have the hardest time arguing around on Aug. 19.