Elon Musk's AI company xAI has filed a lawsuit in U.S. federal court seeking to block a soon-to-be-enacted AI image-editing law in Minnesota. The law, touted as the first state-level regulation specifically targeting AI-powered "de-masking" platforms, is scheduled to take effect on August 1st.
xAI is suing Minnesota Attorney General Keith Ellison. The company claims that the law, ostensibly aimed at combating the generation of nude images without consent, is too broad and could potentially include a large number of protected image expressions in the penalties.
The maximum penalty for each image is $500,000.
The legal number involved is HF 1606. According to the complaint, the platform may be held legally liable as long as its users generate realistic nude images of real people, regardless of whether the platform was aware of it or whether it actively prohibited such behavior.
By law, each infringing image can trigger a civil penalty of up to $500,000. xAI argues that this "strict liability" arrangement does not provide a compliance buffer for general-purpose AI tools.
- Whether the platform was aware of the situation does not affect the establishment of liability.
- Whether the image is disseminated does not affect the establishment of liability.
- Whether the party involved agrees or not does not necessarily constitute a defense.
xAI claims the definition is too broad and lacks exemptions.
In its lawsuit, xAI stated that it does not object to prohibiting the dissemination of AI-generated nude images without consent, but argues that the definition adopted by Minnesota goes beyond this objective. The lawsuit claims that the state government borrowed the "private parts" definition originally used in criminal sexual contact cases and directly applied it to the image generation scenario.
According to xAI, this definition may include content such as the inner thighs, buttocks, or breasts. Therefore, political satire, altered swimsuit photos, or even images of public figures in beach scenes could potentially trigger penalties.
xAI also pointed out that the law does not provide a safe harbor for platforms' well-intentioned compliance measures. Even if a platform has set up a filtering system, the platform may still be held accountable if users bypass the restrictions and generate related images.
The controversy arose after the Grok deepfake scandal.
This lawsuit comes at a time when xAI's image products are under continued pressure. The report mentions that Grok's image tools, launched in July 2025, sparked controversy shortly after being able to generate deepfake images of realistic people using simple prompts, and faced regulatory scrutiny in Europe, Australia, and several US states.
In its lawsuit, xAI disclosed that in 2026, the company suspended over 50,000 accounts and submitted over 70,000 reports to the National Center for Missing and Exploited Children, leading to at least 244 arrests. However, the company argues that these law enforcement collaborations and content governance records do not constitute grounds for exemption from liability under HF 1606.
In contrast, xAI also mentioned the Take It Down Act at the federal level in the United States. This law, signed into law by Trump in May 2025, targets the unauthorized dissemination of private deepfake content, but has narrower application conditions, requiring proof of no consent and actual dissemination, and retaining exceptions for educational and medical content.
Minnesota Attorney General Ellison stated after the lawsuit was filed that AI "disrobing" would deprive victims of their dignity and could cause serious emotional, personal, and professional harm. xAI is currently requesting a federal judge to issue an injunction before the law officially takes effect.
