DOJ Backs xAI’s Challenge to Minnesota’s Fake-Nude Law, Stops Short of Seeking a Court Pause

The Justice Department has entered xAI’s challenge to Minnesota’s new AI “nudification” law with an argument that the state went beyond federal restrictions. But the department did not ask the court to stop the law immediately, leaving xAI’s request for a preliminary injunction with Judge Donovan Frank.

Federal support, limited relief

The DOJ filed a statement of interest in xAI’s lawsuit against Minnesota. Its position is that the state law is broader than existing federal rules and that a growing patchwork of state AI regulations could impede industry development. The filing aligns the federal government with xAI’s challenge to the statute’s scope, without endorsing the company’s specific request to suspend enforcement while the lawsuit proceeds.

Minnesota’s law took effect August 1. It regulates the companies operating websites, apps, software and other services that let users alter an image or video to create fake nudity, rather than targeting individual users. That choice places the compliance burden on the service provider that makes the capability available.

$500,000Maximum civil penalty

Each unlawful access, download, or use can carry a civil penalty of up to $500,000 under Minnesota’s law.

A dispute over breadth and safeguards

xAI sued Minnesota Attorney General Keith Ellison in late July, alleging the law is unconstitutional and violates the First Amendment. The company has argued that the measure is overbroad because it does not expressly account for the user’s intent or whether the person depicted consented to the alteration. At the injunction hearing, xAI also sought carve-outs for consent, artistic expression and parody.

The two approaches now in tension

• Minnesota’s approach is preventative: it prohibits covered service operators from allowing users to “nudify” images or videos.
• The federal Take It Down Act, cited by DOJ, requires platforms to remove reported nonconsensual intimate images, including AI-generated deepfakes, within 48 hours after a valid removal request.
• DOJ says federal rules address many of the same harms more narrowly; Minnesota argues its law can hold providers responsible even if their terms of service prohibit the conduct but users can still generate the images.

The disagreement is not merely about whether nonconsensual sexualized images are harmful. It is about where regulation should attach: after a report triggers removal under the federal framework, or earlier at the point a platform offers the image-alteration capability. Minnesota has cited research saying about 95% of deepfake and manipulated images are nonconsensual, sexually explicit depictions of other people; that figure is the state’s cited evidence in the litigation.

The request remains before the court

Frank had already denied xAI’s request for a temporary restraining order, citing the company’s timing: it sought emergency relief only days before the law was due to take effect. After the later hearing, he took xAI’s preliminary-injunction request under advisement and said he would rule as soon as possible. Until that decision, the legal question is whether xAI can obtain a court-ordered pause while its broader constitutional challenge moves forward.