xAI Sues Minnesota Over $500K-Per-Violation AI Nudification Ban in First Amendment Test

For legal practitioners and regulators, this lawsuit is a landmark First Amendment challenge that will shape the boundaries of AI content moderation. It tests whether a state can penalize platforms up to $500,000 per instance for hosting AI-generated imagery, even when the depicted person consents or the body part definition stretches to publicly visible areas like arms and navels. The outcome will influence not only nudification bans but the broader regulatory toolkit available to states aiming to curb harmful deepfakes.

On the eve of Minnesota's first-in-the-nation ban on 'nudification' technology taking effect, Elon Musk's AI company xAI filed a federal lawsuit challenging the law as an unconstitutional restriction on speech. The suit, filed Monday in the U.S. District Court for the District of Minnesota, marks a critical flashpoint in the longer-running battle over how states may regulate artificial intelligence, particularly when content generated by AI touches on sensitive issues like nonconsensual intimate imagery.

“

It tests whether a state can penalize platforms up to $500,000 per instance for hosting AI-generated imagery, even when the depicted person consents or the body part definition stretches to publicly visible areas like arms and navels.

The law, signed by Governor Tim Walz in May and set to take effect Saturday, August 1, 2026, targets websites and apps that 'knowingly' host or distribute AI-generated nude images of real people without consent. It carries a penalty of up to $500,000 per violation. While xAI does not contest the state's interest in preventing such harm, its 38-page complaint argues the statute sweeps far too broadly, ensnaring a range of constitutionally protected expression. Among its claims: the law prohibits images that were consented to or even self-created by the person depicted; it defines 'intimate part' so expansively that it encompasses body parts routinely visible in public, such as arms or navels; and it provides no safe harbor for platforms that make good-faith efforts to filter violative content.

Minnesota Attorney General Keith Ellison, who had not yet been served at the time of reporting, issued a statement calling the technology 'appalling' and emphasized the profound emotional and professional harm it inflicts on victims. His stance reflects a growing consensus among lawmakers that existing harassment and revenge-porn statutes are insufficient to address the volume and realism of AI-generated deepfakes.

The case arrives as state-level AI regulation proliferates in a federal vacuum. Several states have enacted laws targeting election deepfakes, with mixed results when challenged in court. California's law restricting deceptive media near elections survived a preliminary injunction, while a Texas statute was partially blocked. The Minnesota nudification ban, however, is the first to confront the specific use case of intimate image generation, placing it on untested constitutional terrain.

For xAI, whose chatbot and image generator Grok operates on the X platform, the stakes are immediate. The company argues the law forces it to either over-censor lawful content or face ruinous liability. That dilemma is shared by many AI developers—from startups to tech giants—who increasingly navigate a patchwork of state rules that could shape product design, training data curation, and content moderation systems. A ruling against Minnesota could curtail states' ability to impose content-based restrictions on AI outputs, while a ruling upholding the law might embolden other jurisdictions to pass similar or even more stringent measures.