Access the court order here.
Elon Musk’s artificial intelligence company xAI has sued Minnesota Attorney General Keith Ellison in federal court. The company argues that a state law banning access to AI tools that create fake nude images violates the US Constitution’s free speech protections.
The company filed the case on Monday, days before the law is due to take effect on August 1, 2026. It has asked the court to declare the law unconstitutional and stop Minnesota from enforcing it.
The law requires companies running websites, apps or software to block access to nudification tools. These tools can digitally remove a person’s clothing or create fake nude images. It also bans advertising such services.
A person shown in such an image can sue the company responsible and seek compensation, punitive damages and legal costs. The state attorney general can also seek a civil penalty of up to $500,000 for each unlawful access, download or use.
The case could test how far US states can go in regulating AI products that can be used to create non-consensual sexual images.
What xAI argues: xAI does not challenge Minnesota’s stated aim of preventing non-consensual sexual images. However, it argues that the law goes beyond that purpose and could also cover lawful expression.
According to the lawsuit, the law may also apply to images created with the consent of the person shown. It may also cover images made by that person. The company says it could even apply to content with artistic, humorous or political value.
The company also argues that Minnesota has defined intimate body parts too broadly. It further says the law does not protect companies that make serious efforts to prevent misuse.
The complaint says liability could apply “even if the company has deployed near-perfect, state-of-the-art technical controls to prevent the generation of nude images.”
It also argues that companies could be punished for content created by users who bypass their safeguards.
“Images and videos that users might manage to create in violation of Grok’s terms of service, which is possible only by evading sophisticated, state-of-the-art controls that xAI employs,” the lawsuit states.
xAI says the law will force it to limit Grok Imagine’s image-editing functions. It argues that this could also restrict content protected under the First Amendment.
The company also points to existing state and federal laws. These include the federal TAKE IT DOWN Act. xAI says these laws already address the distribution of non-consensual intimate images. They also require platforms to provide removal mechanisms.
The lawsuit is the latest in a series of legal challenges by Elon Musk’s companies against state AI laws. In 2025, X challenged Minnesota’s election deepfake law on the grounds that it violated the First Amendment and the liability protections available to online platforms under Section 230 of the Communications Decency Act. The company also filed a similar challenge against California’s election deepfake law.
Minnesota defends the law: Ellison said his office had not yet received or reviewed the lawsuit when he issued a statement.
“But I know that using AI to generate nude images of people against their will is appalling. There are plenty of worthy debates to have about AI policy. This is not one of them. AI nudification robs the target of their dignity and can cause immense harm on an emotional, personal, and professional level,” he said.
Minnesota’s law differs from many existing deepfake laws because it targets the companies providing nudification technology, not only the people who use it to create or distribute images.
Texas has taken a narrower approach. Its law can hold a platform liable when it knows that an image was created without consent or fails to remove it after receiving notice.
Grok’s deepfake controversy: The lawsuit follows a global backlash against Grok earlier this year. The criticism centred on its role in generating sexualised images of women and children.
An analysis by the Center for Countering Digital Hate estimated that Grok generated about 3 million sexualised images. The estimate covered an 11-day period between December 29 and January 8. It said around 23,000 of them depicted children.
“That is a shocking rate of one sexualized image of a child every 41 seconds,” the organisation said.
The European Union and the United Kingdom opened investigations, while officials in India, France and Malaysia also raised concerns.
xAI later said it would restrict image editing involving real people in revealing clothing in countries where such content is illegal.
“We have implemented technological measures to prevent the Grok account from allowing the editing of images of real people in revealing clothing such as bikinis, underwear and other revealing attire,” the company said in January.
The company says Grok’s terms prohibit illegal, abusive or privacy-violating use, including the creation of non-consensual nude or sexual images. It says it can suspend users, terminate accounts and report suspected child sexual abuse material.
“xAI takes violations of its terms of service seriously,” the lawsuit states.
The company has also sued a user accused of using Grok to generate child sexual abuse material.
However, the Minnesota case raises a broader question. Can laws targeting AI tool providers survive constitutional scrutiny? The issue is whether laws targeting AI tool providers can survive constitutional scrutiny when those tools can produce both illegal material and constitutionally protected expression.
Read more:
