xAI Loses Bid to Block Minnesota AI Non-Consensual Nude Image Ban; Ban Takes Effect August 1 with $500,000 Penalties

A federal judge refused xAI's request to suspend Minnesota's HF 1606. The law prohibits apps and websites from generating non-consensual intimate images, with the ban set to take effect on August 1, 2026. Violating companies can be fined up to $500,000 per occurrence.

On July 27, 2026, xAI filed suit in the U.S. District Court for the District of Minnesota against Attorney General Keith Ellison, arguing that the law is content-based and overinclusive, encompassing non-sexual, non-consensual, and non-distributed speech protected by the First Amendment. In a ruling on July 31, Judge Donovan Frank noted that xAI waited nearly three months after the law was signed before filing its motion, indicating the alleged harm was not imminent, and therefore denied the application for a temporary restraining order.

How the Law Works

HF 1606 extends regulatory reach to upstream tool providers rather than targeting only users. Minnesota already had criminal laws prohibiting the distribution of non-consensual deepfakes; the new law directly targets AI tools capable of generating or modifying images to depict intimate scenes, regardless of whether developers have deployed content filters or whether harmful intent exists. The Senate passed it 65-0 and the House 132-1, reflecting bipartisan support.

The law provides a path for victims and the Attorney General's office to bring lawsuits against technology that uses social media photos to generate hyperrealistic sexual images and videos. The law's author, State Senator Erin Maye Quade, initiated the legislation two years ago after women reported that their personal photos had been used for such purposes.

Impact on Stakeholders

For xAI, the denial of preliminary relief means its Grok-related image generation capabilities operating in Minnesota must be adjusted immediately, or the company faces penalties of $500,000 per instance. xAI has stated it supports efforts to prevent non-consensual deepfakes but believes the law fails to account for the safety measures it has already implemented.

For other AI developers, companies need to assess whether their tools possess nudification capabilities and redesign or restrict those features in the state. Rising compliance costs may push some companies to exit the Minnesota market or strengthen geofencing measures.

For users and victims, the ban restricts channels for generating non-consensual images through AI tools once it takes effect, and state law enforcement gains new tools to hold companies accountable rather than only individuals. Attorney General Ellison said he will defend the dignity of Minnesotans.

Parallel Precedent

A similar lawsuit in 2025 involving a Musk-affiliated entity that challenged an earlier version of Minnesota's deepfake law was likewise unsuccessful. The state's track record of defending such statutes in court may have influenced the judge's decision to allow HF 1606 to proceed.

Outlook

The case remains at an early stage. Ellison must submit his opposition to the preliminary injunction by August 12, xAI's final brief is due by August 17, and a hearing will be held on August 19. The court's denial of temporary relief indicates it does not view xAI's likelihood of ultimate success as high. If xAI loses again at the preliminary injunction hearing, more states may follow suit in adopting a direct liability model targeting AI tool providers; if xAI advances its constitutional claims at the merits stage, it could trigger conflicts among multiple state regulations.