Judge rejects xAI request to block Minnesota deepfake nudity law

Court denies emergency relief in xAI lawsuit

A federal judge in Minnesota on Monday refused to temporarily block the state's new prohibition on AI-generated nude imagery, handing an early defeat to Elon Musk's artificial intelligence company xAI. The ruling from U.S. District Judge David S. Doty means the law, known as the Minnesota Nudification Ban, will remain in effect while xAI's broader constitutional challenge proceeds through the courts.

The lawsuit, filed earlier this month, argued that the statute is unconstitutionally vague and overbroad, potentially criminalizing legitimate uses of generative AI such as medical research, historical analysis, and satire. xAI sought a preliminary injunction to pause enforcement while the case is litigated, but Doty determined that the company had not demonstrated a likelihood of irreparable harm. In a 23-page opinion, the judge wrote that xAI's concerns were "speculative" and that the statute contains explicit exceptions for lawful conduct.

Background: Minnesota's nudification ban

Minnesota Governor Tim Walz signed the law in May 2025, making the state one of the first to explicitly regulate "deepfake nudification" services. The statute makes it illegal to produce, distribute, or possess AI-generated nude images of a real person without their consent. Violations are punishable as felonies, with enhanced penalties if the depicted individual is a minor or if the image is used for harassment or extortion.

Supporters of the law point to a dramatic rise in nonconsensual intimate imagery (NCII) powered by generative AI. According to the Minnesota Department of Public Safety, reports of AI-generated abuse images tripled in the state between 2023 and 2024. Advocates argue that existing laws, which often require proof that a real nude photo was taken, do not cover synthetic images that are indistinguishable from authentic ones.

"This law closes a critical loophole," said State Senator Erin Murphy, a Democrat who co-sponsored the bill. "For too long, victims of digital impersonation had no legal recourse because the image wasn't 'real.' But the harm to the person depicted is very real — the humiliation, the fear, the damage to their reputation."

The Minnesota law also requires AI platforms to implement age-verification systems and to remove reported deepfake content within 48 hours. Noncompliant companies face fines of up to $50,000 per day.

xAI's legal argument

xAI, which develops the Grok chatbot and image-generation models, claimed in its complaint that the statute violates the First Amendment and the Commerce Clause. The company argued that nudification technology has legitimate applications, including in dermatology, forensic reconstruction, and artistic expression. It also contended that the 48-hour takedown requirement is impossible to meet for open-source models that are distributed outside the company's control.

"The State of Minnesota has enacted a sweeping ban that reaches far beyond the harm it claims to address," said attorney Daniel Petrocelli, who represents xAI. "It imposes strict liability on any developer whose tools could theoretically be misused, without requiring evidence of intent or knowledge. That is a textbook prior restraint on speech."

The company's request for a preliminary injunction was accompanied by declarations from AI researchers who warned that the law could chill innovation. One expert, computer science professor Laks Lakshmanan, testified that the statute's definition of "intimate image" is so broad that it could include anatomical drawings, medical scans, or even a photo of a person in swimwear that is algorithmically modified.

The judge was not persuaded. In his opinion, Doty emphasized that the statute includes an express exclusion for "artistic, educational, or scientific purposes" and that the state's interest in protecting privacy and preventing harassment is paramount. He also noted that xAI does not currently offer a nudification feature to Minnesota users, undermining the claim of imminent harm.

Legal experts weigh in

The ruling is the latest in a series of federal cases testing the limits of AI regulation. Courts have generally upheld laws targeting nonconsensual intimate imagery, but have been more skeptical of statutes that sweep in broad categories of generated content. Legal scholars say the Minnesota case could become a bellwether.

"The judge's denial of injunctive relief is significant, but it is not a final ruling on the merits," said Sarah Chen, a constitutional law professor at Georgetown University. "What matters is that the court engaged with the First Amendment arguments and still found the state's interests compelling. That bodes well for the law, but xAI will have another chance at summary judgment or trial."

Chen added that the "artistic, educational, or scientific" exception is likely to become the central battleground. "The question is whether that exception is broad enough to cover the kind of expressive uses xAI wants to protect. If the court interprets it narrowly, the law could indeed chill some legitimate speech."

Others argue that the judge's ruling was procedural rather than ideological. "A preliminary injunction requires a showing of immediate, irreparable harm. xAI basically said 'we might be sued someday,' and that's not enough," said Marcus Reed, a civil liberties attorney who has followed AI litigation. "The real test comes when the court looks at the statute's vagueness claim in detail."

Implications for the AI industry

The decision could have ripple effects far beyond Minnesota. Several other states, including California, New York, and Illinois, have introduced or passed similar measures in response to the surge in deepfake abuse. If Minnesota's law survives judicial scrutiny, it may serve as a template for federal legislation now pending in Congress.

For AI companies, the ruling signals that courts are unwilling to grant broad exemptions from deepfake regulations based on hypothetical future harms. That could force developers to build stricter content filters, watermarking, and opt-in consent mechanisms into their products — changes that xAI and others have resisted as costly and technically difficult.

"This is a wake-up call for the industry," said Anita Deshpande, a policy analyst at the Center for Digital Ethics. "You can no longer argue that your model is a neutral tool when it is designed to remove clothing from photos. The law is catching up with the technology, and companies need to adapt."

xAI has not indicated whether it will appeal the injunction ruling. The company's core lawsuit is still pending, and a trial date has not been set. In the meantime, Grok's image-generation feature is unavailable to users with Minnesota IP addresses, a measure xAI implemented voluntarily last week.

What happens next

The case will now proceed to discovery, during which both sides will gather evidence on the statute's scope and the technical feasibility of compliance. Legal analysts expect xAI to file motions to dismiss or for summary judgment within the next few months. If the case reaches trial, it could become one of the first appellate decisions on the constitutionality of deepfake-specific laws.

Minnesota officials, meanwhile, have welcomed the ruling. Attorney General Keith Ellison's office issued a statement saying the state "remains committed to protecting its citizens from nonconsensual synthetic imagery." The Department of Public Safety has begun training local law enforcement agencies to enforce the new law, and prosecutors are preparing model cases to test its boundaries.

For advocates, the ruling is a crucial validation of a victim-centered approach. "Today's decision sends a clear message that privacy outweighs the convenience of AI novelty," said Maya Torres, executive director of the Digital Safety Coalition, a nonpartisan group that supported the law. "We hope other states follow Minnesota's lead."

Observers note that the broader legal landscape is still uncertain. The U.S. Supreme Court has yet to rule on any case involving AI-generated content, and federal efforts to criminalize nonconsensual deepfakes remain stalled in committee. Until then, state laws like Minnesota's will serve as the primary deterrent — and the primary battleground for tech companies that oppose them.

xAI's lawyers have declined to comment on the ruling beyond a brief statement: "We are reviewing the order and evaluating our options." The company has filed no notice of appeal as of press time.

A shifting regulatory landscape

The Minnesota case is unfolding against a backdrop of global action. The European Union's AI Act, which took full effect in August 2025, imposes transparency and mitigation requirements on generative models. The United Kingdom is considering a proposed offense of "creating a deepfake of a person without consent." And in Asia, South Korea and Japan have already enacted stricter penalties for synthetic sex crimes.

Industry analysts say the patchwork of regulations creates compliance headaches for AI companies that operate globally. "A model trained in San Francisco, served from Oregon, and available to a user in Minnesota may be subject to different rules in each jurisdiction," said technology lawyer James Okafor. "The xAI lawsuit is an attempt to get a court to impose a uniform national standard — but the judge's ruling suggests that attempt will not succeed easily."

The next substantive hearing in the case is scheduled for October 14, when the court will hear arguments on xAI's motion to dismiss. Until then, Minnesota's nudification ban remains in full force.

Conclusion

Judge Doty's denial of xAI's injunction request allows Minnesota to enforce one of the nation's most aggressive deepfake laws, at least for now. While the final constitutional verdict is far from settled, the ruling establishes an early marker: courts are likely to treat the protection of human dignity and privacy as compelling government interests, even when weighed against the expressive potential of AI tools. For xAI and other developers, the path forward is no longer about whether to comply with state regulations, but how to build products that can coexist with them. For victims of image abuse, the decision offers a measure of hope that the law will shield them from the next generation of exploitation.