Elon Musk's company xAI has filed a lawsuit against Minnesota, challenging the state's new law that bans the use of artificial intelligence to create fake nude images of real people. The lawsuit, filed in federal court, contends that Minnesota's first-in-the-nation legislation is unconstitutional and overly broad.
The law, which was set to take effect Saturday, aims to outlaw the rapidly spreading technology that allows individuals to generate non-consensual nude images of others using AI. xAI, whose AI model Grok is available on the social media platform X (formerly Twitter), stated in its lawsuit that it does not object to the state's goal of preventing the distribution of AI-generated nude images without consent. However, the company argues that the Minnesota law extends beyond this objective, infringing upon constitutionally protected speech and potentially penalizing the company with $500,000 per violation.
xAI's legal challenge asserts that the law lacks a "safe harbor" provision for companies making good-faith efforts to prevent misuse of their AI tools. The lawsuit also claims the law covers images where consent was given by the depicted individual or even created by that person. Furthermore, xAI argues that the law's definition of "intimate part" is too broad, encompassing body parts frequently visible in public.
Minnesota Attorney General Keith Ellison stated that his office had not yet been served with the lawsuit. He commented that using AI to create nude images of individuals against their will is "appalling" and causes significant harm. Ellison acknowledged that there are valid debates surrounding AI policy but believes this specific issue is not one of them.
This legal challenge comes as states grapple with regulating AI-generated content. Previous attempts at similar laws, such as California's legislation on election deepfakes, have faced challenges in court over First Amendment protections. Minnesota's law, however, had reportedly survived an earlier legal challenge from Musk's company prior to its enactment.
The Minnesota law differs from existing state and federal statutes that criminalize the creation and distribution of non-consensual sexually explicit deepfakes. Those laws typically target the users generating the images rather than the developers of the AI tools themselves. The federal Take It Down Act, passed in the previous year, mandates that websites and apps establish processes for removing such content.
In its lawsuit, xAI highlighted its own terms of service, which prohibit the use of Grok for illegal, harmful, or abusive activities, including altering images to create non-consensual nudity or sexually explicit depictions. The company stated it enforces these rules through account suspensions or terminations and by reporting suspected child sexual abuse material.
Earlier this year, xAI declared a "zero tolerance" policy for child sexual exploitation, non-consensual nudity, and unwanted sexual content. Following global criticism regarding the sexualization of images of women and children, Grok announced in January that it would implement geo-blocking to prevent content violations in specific regions. The company also stated it had deployed technological measures to stop users from editing images of real individuals in revealing attire, such as bikinis or underwear, where such edits are illegal.