Elon Musk’s xAI Sues Minnesota Over AI Nudification Ban

Elon Musk’s artificial intelligence company xAI has sued Minnesota over the state’s first-in-the-nation ban on tools that create realistic nude images of identifiable people, arguing the law restricts protected expression along with abusive content.

The company filed the case July 27 in federal court, days before the measure took effect Aug. 1. xAI operates Grok, an AI chatbot and image generator available through Musk’s social media platform X and other services. The company is asking a judge to declare the law unconstitutional and block Minnesota Attorney General Keith Ellison from enforcing it against xAI.

Minnesota’s law prohibits website, app and software operators from allowing users to access technology that can “nudify” an image or video. Companies also may not create those images for users or promote services that perform the same function.

An image falls under the statute when technology adds or generates an intimate body part that was not shown in the original and produces a result realistic enough to appear to depict the identifiable person. The attorney general may seek penalties of up to $500,000 for each unlawful access, download or use.

People shown in prohibited images may also sue for damages, attorney fees and a court order stopping the conduct.

Minnesota’s approach differs from many laws aimed at sexually explicit deepfakes. Those measures often punish the person who knowingly creates or shares a nonconsensual image, while federal law also requires covered platforms to remove reported material. Minnesota places responsibility on the company providing the tool before an image is distributed.

xAI says it does not dispute the state’s authority to protect people from nude images made or shared without their consent. The complaint claims, however, that Minnesota’s law extends beyond that goal because it does not require proof that the person shown objected to the image or that anyone distributed it.

The ban could therefore apply when adults alter their own photographs, agree to appear in generated images or create material that remains private, according to the filing. xAI also argues that the state’s definition of an intimate body part is broad enough to cover areas commonly visible in public, including in swimwear.

The constitutional challenge turns on whether Minnesota wrote the law narrowly enough to address the harm it seeks to prevent. Restrictions based on the content of speech are generally reviewed under strict scrutiny, one of the Constitution’s most demanding legal tests.

To survive that review, Minnesota would need to show that the law serves a compelling public interest and restricts no more expression than necessary. Preventing nonconsensual sexual images may qualify as such an interest, but xAI argues the statute also reaches consensual, private, artistic, political, educational, medical and satirical material.

The company also raises a First Amendment doctrine known as overbreadth. Courts may strike down a speech restriction when it covers a substantial amount of lawful expression along with conduct the government may regulate.

According to the complaint, Minnesota could have adopted a narrower measure focused on whether a person consented, whether the creator knew consent was lacking and whether the image was distributed. The current law does not include those limits.

xAI separately objects to the absence of an intent or knowledge requirement. State officials would not have to show that an AI company meant to produce a prohibited image or knew that a user was attempting to create one.

No safe harbor protects companies that make good-faith efforts to prevent misuse. As a result, xAI claims a provider could face liability when a user defeats its safeguards, even when the company banned the conduct and tried to stop it.

Grok’s terms prohibit users from altering a real person’s image to show nudity or sexual content without consent, the company states. xAI says it may suspend or terminate accounts for violations and report suspected child sexual abuse material.

Those policies do not decide whether the company has complied with Minnesota law. They support xAI’s argument that the statute could punish a provider despite efforts to prevent prohibited images.

The law includes an exemption for services that require substantial human technical or artistic skill and judgment. xAI claims that distinction burdens widely available AI tools while allowing skilled users to produce similar images through more complex software.

Ellison has defended the law’s purpose, saying AI-generated nude images made against a person’s will can cause serious emotional, personal and professional harm. His office had not reviewed the complaint when the Associated Press requested a response and had not yet filed its full answer to xAI’s constitutional arguments.

U.S. District Judge Donovan Frank denied xAI’s request for a temporary restraining order on July 31, finding that the company’s delay in seeking emergency relief weighed against stopping the law before it took effect. The ruling did not decide whether the statute violates the First Amendment.

Minnesota must respond to xAI’s request for a preliminary injunctionpreliminary injunctionA preliminary injunction is a temporary court order issued before a case is finally decided that requires a party to take or stop an action while the litigation continues. Federal courts generally consider the likelihood of success, irreparable harm, the balance of equities, and the public interest before granting one.Read more → by Aug. 12. The company’s reply is due Aug. 17, followed by an Aug. 19 hearing in St. Paul.