A remarkable new law went into effect in Minnesota on Saturday that bans nudification apps, AI-based or not. It penalizes the app provider by enabling lawsuits from victims, and levying fairly harsh fines of up to $500,000 not just per generation, but every time someone accesses the image, meaning a viral image could be financially catastrophic.
Just before it went into effect, District Court Judge Donovan Frank denied a last minute legal request from xAI, sometimes known as SpaceXAI, the AI division of the Elon Musk-led conglomerate, and the originator of the Grok AI chatbot. XAI’s request for a temporary restraining order would have effectively blocked enforcement of the law.
The law defines “nudification” as altering a photo of a recognizable person to expose “intimate parts,” which—with apologies for being a bit graphic—can include more than just the classic broadcast TV standard of genitals, buttocks, and female nipples. The “intimate parts definition comes from a separate state statute defining those parts as the “primary genital area, groin, inner thigh, buttocks, or breast of a human being.”
To count as nudification, image alteration has to be basically at-the-push-of-a-button, and can’t involve artistry or skill. So the pervy Photoshoppers of Minnesota are safe from this law (but this is not legal advice, and don’t come crying to Gizmodo if you run afoul of Minnesota’s revenge porn laws).
SpaceXAI’s suit against the Minnesota Attorney General takes issue with the breadth of the law, which, it argues, is expansive enough to include many AI image generations created in ostensibly innocent fun. In particular it cites AI generated images of Trump Administration officials in bathing suits. Another image included in the suit (in full color) features former New Jersey Governor Chris Christie and current Illinois Governor J.D. Pritzker squaring off against one another while dressed only in what are apparently supposed to be sumo mawashi.
At least part of Judge Frank’s legal reasoning was timing. “xAI filed the motion on July 29, 2026, nearly three months after the law was signed, and only three days before the law is set to take effect,” he wrote, adding, “Such a delay in bringing the action and the motion suggests that harm is not immediate.”
The Grok AI chatbot, which at the time was the flagship XAI product, became a nasty meme right around the most recent holiday season by non-consensually generating sexualized and/or scantily-clad images of people. According to one estimate, Grok created 23,000 such images of children. In a statement posted to X, which is the SpaceX-owned social media outpost of the Grok chatbot, X said the capabilities that enabled this behavior had been curtailed. The statement also provided assurance that X’s operators “continue to have zero tolerance for any forms of child sexual exploitation, non-consensual nudity, and unwanted sexual content.”