A federal judge has denied xAI's request for a temporary restraining order against Minnesota's first-in-the-nation ban on AI nudification apps, allowing the law to take effect on August 1, 2026. The ruling from US District Judge Donovan Frank leaves the law in force while xAI's broader constitutional challenge continues through the courts. NBC News

The law prohibits apps and websites that can generate nonconsensual intimate imagery of identifiable individuals. Minnesota defines a "nudified" image or video as one "altered or generated to depict an intimate part not depicted in an original unaltered image or video of an identifiable individual," where the result is "so realistic that a reasonable person would believe that the intimate part belongs to the identifiable individual." Developers face fines of $500,000 each time a user generates a nonconsensual intimate image using their products. Engadget

Judge Frank was not persuaded that enforcement would cause immediate harm. He noted that xAI filed its lawsuit on July 29, 2026, nearly three months after the law was signed and only three days before its effective date, which undercut the urgency argument. A hearing on whether to grant xAI's request for a preliminary injunction (a court order that would pause the law while the case proceeds) is scheduled for August 19. Engadget

xAI, owned by SpaceX as of July 2026, sued Minnesota Attorney General Keith Ellison to challenge the statute. The company argues the law is an "overbroad, content-based ban on free speech and the tools of visual expression" that extends beyond preventing nonconsensual deepfake imagery and exposes protected speech to civil liability and government sanction. CNBC

Despite the denial, xAI's underlying lawsuit remains active. The restraining order was a stopgap measure; the preliminary injunction hearing and the merits of the First Amendment challenge are still ahead. Engadget

Grok, xAI's chatbot, implemented policy changes by mid-January to address requests that transform photos of real women and children into sexualized images. NBC News reported that Grok was still undressing real people as recently as April. Engadget

The broader context here matters for anyone building generative image models. Minnesota's $500,000-per-incident penalty structure creates a liability exposure that is not theoretical. If a model provider's product can be used to generate a realistic nude of an identifiable person, each such generation is a separately fineable event under this statute. That is a different enforcement mechanism than criminal penalties against individual users, which several states have already enacted. It targets the developer directly.

xAI's First Amendment argument follows a familiar pattern in technology regulation: the company is not challenging the goal of preventing nonconsensual intimate imagery but the breadth of the mechanism. The claim that a ban on "tools of visual expression" sweeps in protected speech has precedent in challenges to restrictions on encryption, peer-to-peer software, and other general-purpose technologies. Whether that argument succeeds will likely turn on whether the court accepts the law as narrowly tailored to the harm it addresses or sees it as a content-based restriction that fails strict scrutiny, the highest standard of judicial review applied to laws that regulate speech based on its content.

The August 19 preliminary injunction hearing is the next concrete milestone. If Judge Frank declines to issue the injunction at that stage, the law remains in force through the duration of the litigation, and Minnesota's enforcement apparatus can begin operating in earnest. Other states are watching. Minnesota is described as the first-in-the-nation law of this kind, but state-level AI regulation has tended to cascade quickly once one jurisdiction establishes a workable framework.

For model providers whose systems can generate or alter photorealistic human bodies, the practical question is whether existing safety filters and policy guardrails are sufficient to prevent the specific outputs this law targets. Grok's gap between announced policy changes in January and continued generation of sexualized images in April illustrates the difficulty. Technical safeguards against nonconsensual nudification are not a solved problem, and a half-million-dollar-per-incident liability makes the engineering investment to solve it a directly quantifiable business decision rather than a reputational one.