Elon Musk’s xAI has filed a lawsuit against the state of Minnesota regarding its landmark legislation that prohibits “nudification” technology on digital platforms, marking a significant legal challenge about the extent of state powers in regulating artificial intelligence under constitutional principles.
The lawsuit was initiated in federal court on Monday, just a few days before the law is scheduled to become active on Saturday. If enacted, Minnesota will be the first state to attempt to ban the burgeoning technology that allows users to generate AI-driven fake nude images of individuals. This law, which was signed in May, has attracted considerable attention.
In the 38-page filing, xAI, which offers the AI chatbot and image generator Grok on the social media platform X (previously Twitter) and other venues, expressed that it does not dispute the state’s aim to prohibit the unauthorized dissemination of AI-generated nude representations of individuals. However, the lawsuit alleges that the new law “overreaches,” prohibiting many images and videos that are protected under constitutional rights and imposing penalties of up to $500,000 for each violation.
The company points out that there is no provision for companies trying to avoid creating these types of images through good-faith efforts, as well as claiming that the law includes images that are consented to by the subjects or even created by those individuals themselves.
Additionally, xAI contests the law’s broad definition of “intimate part,” which encompasses body parts commonly seen in public.
Minnesota Attorney General Keith Ellison commented on the lawsuit, stating that his office has not yet been officially notified or received the complaint. However, he noted, “The use of AI to produce nude images without consent is reprehensible. There are many valid discussions to be had regarding AI policy, but this isn't one. AI nudification strips individuals of their dignity and can inflict profound emotional, personal, and professional harm.”
Legal precedents on state regulations concerning election-related deepfakes have yielded mixed judicial outcomes. For instance, California’s analogous law was blocked for violating First Amendment rights, while Minnesota’s legislation has withstood a prior court challenge from Musk’s company in 2025.
This new Minnesota law is distinct from existing state and federal measures that criminalize deepfakes—AI-generated explicit images created without the individuals' consent. Such laws generally target the users of the technology rather than its developers. Last year, the federal Take It Down Act mandated that online platforms establish a process for removing such images.
Texas also approached nudification regulations differently last year, establishing that website or app owners could face civil liabilities and possible criminal charges if they were aware that a depicted individual had not consented and failed to act when notified.
Within its lawsuit, xAI emphasized that Grok's terms of service explicitly forbid usage for “any illegal, harmful or abusive activities” infringing on privacy, including modifying images to create nudified or sexually explicit portrayals. It maintains enforcement of these policies through suspending or terminating accounts and reporting suspected child sexual abuse content.
Earlier in the year, xAI reaffirmed its commitment to a strict stance against child sexual exploitation, non-consensual nudity, and unwanted sexual material, declaring a “zero tolerance” policy.
In response to global pushback over sexualized imagery involving women and children, Grok announced in January that it would implement geo-blocking to restrict content that contravenes local laws, effectively banning edits of images depicting real individuals in revealing outfits where that is prohibited. The platform's “spicy mode” feature, which previously enabled users to craft explicit content, faced widespread criticism from governments around the world.
“Technological measures have been put in place to ensure Grok accounts do not allow edits of real individuals in revealing clothing such as bikinis or underwear,” xAI stated in a January release.
