xAI is trying to stop the xAI Minnesota nudification law fight from becoming an expensive product problem. The company sued Minnesota Attorney General Keith Ellison over a statute passed in May and scheduled to take effect on August 1, arguing in its complaint that the law would leave xAI with “no practical choice” but to limit Grok Imagine’s image-editing tools.
The company’s core claim is constitutional: xAI says Minnesota’s law violates the First Amendment because it reaches too much protected expression while threatening severe penalties. The complaint challenges provisions that target “nudification” technology, a category of tools used to alter images or video so a person appears nude.
The law bars site owners from letting users “access, download, or use” covered nudification software, according to the Minnesota bill text. It includes an exception for technology that requires a user’s technical skill to nudify an image or video. The statute also lets private people sue and authorizes the attorney general to seek penalties of up to $500,000 per violation.
What is Minnesota’s nudification law?
Minnesota’s law is aimed at services that make it easy to generate nude or sexualized images of real people without doing much technical work. In plain English, it tries to hit the app-store version of sexual harassment software rather than every image editor on a laptop.
xAI says that distinction is still too broad. In its complaint, the company argues the law can impose liability for consensual images, adult images that are sexualized but not illegal, and material with humorous or artistic value. It also objects to what it describes as strict liability for users who evade Grok’s terms of service and technical controls.
The complaint says xAI uses “sophisticated” safeguards and takes terms-of-service violations seriously. It points to xAI’s separate lawsuit against a user accused of using Grok to generate child sexual abuse material. That lawsuit was filed more than two months after Minnesota passed the nudification statute.
The legal fight follows Grok’s earlier deepfake disaster. The New York Times reported in January that Grok produced millions of sexually explicit deepfakes, including images involving minors. The Verge also reported that Grok generated images of children and that governments in the European Union and the United Kingdom opened inquiries, while officials in France, India, and Malaysia criticized the output.
The Center for Countering Digital Hate said it examined an 11-day period from December 29 to January 8 and found roughly 3 million sexualized images generated by Grok. The group said about 23,000 depicted children, a rate it described as one sexualized child image every 41 seconds.
xAI’s lawyers argue that Minnesota and federal law already address nonconsensual distribution of nude deepfakes, including statutes such as the TAKE IT DOWN Act. Their position is that those laws are more carefully written, while Minnesota’s statute burdens speech based on content and fails the narrow tailoring required for such rules.
That is the part a court will have to take seriously, even if the product history is ugly. The First Amendment does not disappear because a company shipped a reckless AI feature. It also does not answer the practical question Minnesota is raising: whether existing laws can deter platforms when their tools can generate sexualized deepfakes at industrial scale.
This story draws on original reporting from The Verge.
