Summary
- xAI has initiated a First Amendment lawsuit against Minnesota Attorney General Keith Ellison to contest HF 1606, the nation's first legislation addressing AI nudification platforms, set to take effect on August 1.
- The legislation imposes severe civil penalties of up to $500,000 per image without offering safe harbor, even for platforms that actively work to prevent nudification.
- The law was passed overwhelmingly in Minnesota, with a House vote of 132-1 and a Senate vote of 65-0, following an incident where a man used social media photos to create sexual images of over 80 women.
xAI, the artificial intelligence firm founded by Elon Musk, filed a lawsuit on Monday against Minnesota's Attorney General Keith Ellison, seeking to halt HF 1606, which targets AI "nudification"—software that alters clothing in real images. The company argues that the law constitutes an excessive and content-based restriction on free speech and visual expression in a misguided effort to ban nudification.
The statute, HF 1606, holds any platform strictly liable for user-generated images that depict real individuals in compromising ways, regardless of the platform's intent or knowledge. Violations can result in hefty fines of up to $500,000 per image.
In its complaint filed in the U.S. District Court for the District of Minnesota, xAI does not dispute Minnesota's right to prohibit the unauthorized distribution of nude images but contends that the law's scope is excessively broad, threatening a variety of protected speech with civil penalties and government action. The lawsuit points out that the state's definition of "intimate part" is drawn from a criminal statute intended for nonconsensual touching, which complicates the issue of image generation.
xAI highlights that generating a realistic AI image of a public figure in a swimsuit or a shirtless politician could invoke penalties under this law, even if the images are intended for satire.
The complaint emphasizes the lack of any safe harbor provisions, stating that platforms cannot escape liability even if users consented or created the images themselves. Moreover, the law does not require proof of knowledge regarding wrongdoing, meaning a company could be held liable even if a user bypasses its filters.
In 2026, xAI suspended over 50,000 accounts and reported more than 70,000 instances of potential abuse to the National Center for Missing & Exploited Children, resulting in at least 244 arrests. However, these actions do not serve as a defense under HF 1606.
The broader context reveals that xAI's Grok image tools faced scrutiny shortly after their launch in July 2025 for producing deepfakes—AI-generated images of real individuals—leading to regulatory inquiries in multiple regions. Cities such as Baltimore and three minors in Tennessee have separately sued xAI over Grok's outputs. The company also challenged a Colorado AI law in April on similar First Amendment grounds.
The law HF 1606 passed with overwhelming support in Minnesota after a man misused social media images to create sexual representations of numerous women. Senator Erin Maye Quade, a key supporter of the bill, confirmed that the law intentionally includes consensual images in its prohibitions.
Additionally, the federal TAKE IT DOWN Act, enacted by President Donald Trump in May 2025, already addresses nonconsensual intimate deepfakes but requires proof of nonconsent and actual distribution, applying a narrower definition of body parts and allowing exceptions for educational and medical content. xAI argues that Minnesota should have followed this model instead.
Attorney General Ellison remains firm in his stance, stating, "AI nudification robs the target of their dignity and can cause immense harm on an emotional, personal, and professional level," in response to the lawsuit.
With HF 1606 set to take effect on Saturday, xAI is seeking a federal judge's intervention to prevent its implementation before the deadline.