District Judge Donovan Frank ruled on Friday that the law would take effect as scheduled, noting that xAI’s decision to wait nearly three months after the bill was signed to file its motion suggested there was no risk of immediate harm.
The company had sought to pause the legislation, arguing that it violates the First Amendment’s guarantee of free speech.
The legal challenge is significant because it marks the first major court battle over state-level efforts to regulate AI-generated sexual deepfakes—realistic images or videos altered to show someone naked without their consent.
Under the new law, owners of software or websites can face civil penalties of up to $500,000 per violation for allowing users to generate such content.
While xAI argues that its tool, Grok, already prohibits non-consensual sexual imagery, the company claims the law is overinclusive and could punish legitimate forms of expression, such as satire or political memes.
The court has scheduled a follow-up hearing for August 19 to determine whether to grant a preliminary injunction, which would temporarily stop the law's enforcement while the lawsuit proceeds.
This case affects all AI developers and platform owners operating in Minnesota, as the law does not provide a "safe harbor"—a legal protection for companies that make good-faith technical efforts to prevent abuse.
The ruling comes amid broader national pressure on AI infrastructure providers and app stores to remove tools capable of creating sexualized deepfakes.