Minnesota’s first-in-the-nation prohibition on automated “nudification” services has taken effect after a federal judge refused xAI’s last-minute request to stop the law.
Elon Musk’s AI company argued that the measure threatens constitutionally protected expression and could expose Grok’s image-generation service to enormous financial penalties. Minnesota responded that an emergency order was unwarranted because xAI waited almost three months after the law was signed before seeking relief.
US District Judge Donovan Frank denied the temporary restraining order on July 31, allowing the legislation to become enforceable on August 1. The ruling does not decide whether Minnesota’s law is constitutional, and xAI’s broader lawsuit remains active. A hearing on the company’s request for a preliminary injunction is scheduled for August 19.
The Judge Did Not Rule on the Law’s Constitutionality
The immediate ruling was based primarily on timing rather than a final assessment of xAI’s First Amendment arguments.
According to the court’s two-page order denying emergency relief, xAI filed its motion on July 29—only three days before the law was due to take effect and nearly three months after it had been signed. Judge Frank said that delay suggested the alleged harm was not sufficiently immediate to justify a temporary restraining order before August 1.
The court will instead consider xAI’s request as a motion for a preliminary injunction. Minnesota must file its response by August 12, xAI can reply by August 17, and the hearing will take place in St. Paul on August 19.
That distinction is important. The judge has not concluded that the law is constitutional, narrowly written or certain to survive. He has only refused to suspend it through an emergency order issued immediately before its effective date.
What Minnesota’s New Law Actually Prohibits
Minnesota’s law targets websites, applications, software and other services that allow users to create realistic fake nude images or videos of identifiable people.
The official text of Minnesota Session Law Chapter 72 defines “nudification” as altering or generating media to show an intimate body part that did not appear in the original image, provided the result is realistic enough for a reasonable person to believe that it belongs to the identifiable individual.
A person controlling a covered service may not allow users to access, download or use it for nudification, perform the alteration for a user, or advertise a service that offers the capability. The law includes an exemption for tools that require substantial individual technological or artistic skill from the user, potentially distinguishing automated one-click services from professional editing software.
The wording does not expressly limit the prohibition to images produced without the depicted person’s consent. That omission is central to xAI’s argument that Minnesota has regulated more speech than necessary to address nonconsensual intimate imagery.
Penalties Could Reach $500,000 Per Violation
The law gives the Minnesota attorney general authority to seek civil penalties of up to $500,000 for each unlawful access, download or use of prohibited technology.
A person shown in an illegally nudified image can also bring a private lawsuit seeking compensation, including up to three times the actual damages, punitive damages, an injunction and reasonable legal costs. Money collected through state penalties is intended to support organisations serving victims of sexual assault, domestic violence, child abuse and other crimes.
The scale of those penalties explains why xAI sought emergency protection. An AI service with a large user base could theoretically face extensive exposure when numerous people access or use a prohibited feature.
The law states that it does not modify the protections and liabilities created by Section 230 of federal law. However, the dispute may still test whether a state can directly regulate the functionality supplied by an AI developer rather than focusing only on content later posted by its users.
xAI Says the Ban Is Too Broad
xAI does not dispute Minnesota’s interest in stopping the distribution of nonconsensual synthetic nude images. Its lawsuit instead argues that the state has gone far beyond that objective.
The company says the law could cover consensually generated images, media created by the person depicted and artistic, humorous or political works protected by the First Amendment. It also objects that developers can face liability even when they have made good-faith efforts to prevent misuse.
xAI has described the measure as an overbroad, content-based restriction on speech and visual-expression tools. The company says it may have to restrict Grok Imagine’s image-editing functions for users in Minnesota to reduce its legal risk.
Its challenge may therefore turn on whether the court sees the legislation primarily as regulation of harmful commercial technology or as a restriction on the production of expressive content. Laws aimed at particular categories of images can face demanding First Amendment scrutiny, particularly when their definitions extend beyond clearly unlawful material.
Minnesota Says the Technology Causes Immediate Harm
Minnesota Attorney General Keith Ellison has defended the legislation as necessary protection against tools that can transform ordinary photographs into realistic sexualised images without the subject’s cooperation.
In his statement following the court decision, Ellison said nudification services had been used to harass people and generate child sexual abuse material. He argued that the state has an urgent interest in protecting the dignity and safety of people whose images can be manipulated within seconds.
The law received unusually broad political support. Minnesota’s attorney general said 197 lawmakers voted for it across the state House and Senate, with only one opposing vote. The House passed its version 132–1 before it was presented to Governor Tim Walz and signed on May 7.
Supporters argue that existing laws often punish the person who distributes an abusive deepfake only after the image has already been created and circulated. Minnesota’s approach intervenes earlier by targeting companies that make automated nudification capability directly available.
The Dispute Goes Beyond Grok
The lawsuit has immediate consequences for xAI, but its significance extends to the wider generative-AI industry.
Many image models are general-purpose tools rather than products explicitly marketed as “nudify” applications. The same model may create legitimate artwork, edit clothing, modify photographs or generate abusive sexual deepfakes depending on the user’s instructions and the safeguards built into the service.
Minnesota’s technical-skill exemption also raises difficult boundary questions. A one-click undressing application appears to fall clearly within the law, while complex manual editing software may qualify for the exemption. General-purpose AI systems operating between those extremes may struggle to determine precisely which capabilities must be blocked.
A 2026 safety audit of 155 eligible mobile face-swapping applications found that 70% lacked technical safeguards preventing users from producing nude face swaps. The study concluded that services not openly marketed as nudification tools could still be used to generate synthetic nonconsensual intimate imagery.
That broader availability helps explain Minnesota’s decision to regulate functionality rather than relying entirely on how an application describes itself.
Federal and State Laws Take Different Approaches
The federal TAKE IT DOWN Act and several state laws already address nonconsensual intimate deepfakes, but many focus on distribution, removal procedures or conduct by the person creating and sharing the image.
Minnesota’s measure is different because it targets owners and operators of the underlying technology. Texas has adopted a more conditional approach, creating potential liability when a service provider knows that consent was absent or fails to remove an image after receiving notice. Minnesota’s law does not contain the same knowledge requirement or explicit good-faith safe harbour.
The lawsuit could consequently influence how other states draft similar legislation. A victory for Minnesota may encourage broader bans on automated nudification functions. A successful xAI challenge could push lawmakers toward narrower laws centred on nonconsent, intentional misconduct and notice-based liability.
The August 19 Hearing Will Be the Next Major Test
The law is now enforceable, but its future remains uncertain.
At the preliminary-injunction hearing, xAI will seek to show that it is likely to succeed on the merits, that enforcement would cause irreparable harm and that pausing the law would serve the public interest. Minnesota will argue that the legislation addresses serious and immediate harms while remaining consistent with constitutional and federal protections.
The court could leave the statute in effect, temporarily suspend all or part of it, or interpret its language in a way that narrows how it can be enforced against xAI.
For now, Minnesota has won the first procedural round. Its victory is limited but consequential: automated nudification services face legal exposure in the state while the constitutional fight continues.
The larger question is whether governments can prohibit an AI capability before it is abused or whether they must wait until a specific person creates and distributes an unlawful image. The answer could shape not only Grok’s future in Minnesota, but the emerging legal limits placed on generative image tools across the United States.
