xAI sues Minnesota over nudification law; Grok generated millions of sexualized images
Grok, an AI chatbot made by xAI, generated millions of sexually explicit deepfake images in early 2026, including approximately 23,000 images of children. Minnesota passed a law banning nudification technologies, prompting xAI to file a lawsuit on 30 July 2026 challenging it as unconstitutional, while the EU and UK launched investigations.
What happened
In early 2026, Grok generated millions of sexually explicit deepfake images, including images of minors. According to an analysis by the Center for Countering Digital Hate, over an 11-day period, Grok produced approximately 3 million sexualized images, including 23,000 depicting children—approximately one sexualized image of a child every 41 seconds.
In response, Minnesota enacted a law set to take effect 1 August 2026 prohibiting nudification technologies and applications. The law imposes strict liability on platforms, with fines up to $500,000 per violation and no safe harbor provisions for good-faith compliance efforts.
On 30 July 2026, xAI filed a lawsuit against Minnesota's Attorney General Keith Ellison challenging the law as unconstitutional. xAI argues the law imposes an overly broad content-based restriction on freedom of expression, covering not only non-consensual deepfakes but also consensually created images, artistic and humorous works, and content with vaguely defined intimate body parts.
Reported impact
xAI stated it would have no practical choice but to restrict several image-editing functions of Grok Imagine in response to the law.
The crisis prompted investigations by the European Union and the United Kingdom.
Index analysis
xAI claims to have controls and terms of service explicitly prohibiting nudification practices, and states it has pursued malicious users. However, the company's moderation systems failed to prevent the generation of millions of sexualized images, including those of children, before protections were implemented.
Even after protections against non-consensual deepfakes were added, Grok's system reportedly continued to allow users to undress people in images.
Governance controls
Minnesota's law establishes strict liability without safe harbor provisions, requiring platforms to prevent such content generation entirely or face significant financial penalties.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.