Minnesota can continue enforcing its new law aimed at AI-generated fake nude images while Elon Musk’s xAI challenges it in federal court. The immediate dispute is not over a final constitutional ruling: Judge Donovan W. Frank denied xAI’s request for a preliminary injunction, leaving the statute in effect during the litigation.
2
11
What Minnesota’s law prohibits
Minnesota House File 1606 took effect on August 1, 2026. It targets providers that allow users to use covered AI technology to generate certain sexual images or images of intimate parts involving identifiable people, including through tools such as xAI’s Grok Imagine.
1
9
11
The law is commonly described as a ban on AI “nudification”: using AI to alter or generate a realistic image or video so that an identifiable person appears to have an intimate body part not present in the original material. Reports on the litigation describe the measure as the first such state law in the United States.
2
4
The Minnesota attorney general may seek injunctions and civil penalties of up to $500,000 per violation.
5
9
xAI’s First Amendment case
xAI sued Minnesota Attorney General Keith Ellison in late July, arguing that HF 1606 is an unconstitutional, content-based restriction on speech and visual-expression tools. The company’s filings argued that liability depends on what an edited image depicts and that the law could sweep beyond nonconsensual sexual imagery to protected expression.
26
27
The company sought emergency relief before the law took effect, then a preliminary injunction that would have suspended enforcement while the constitutional case proceeded. xAI also said the law would affect Grok and Grok Imagine users in Minnesota.
32
36
Minnesota disputed that framing. Ellison argued that the law is narrowly tailored to address nonconsensual AI sexual imagery and that the challenged feature is a product capability subject to regulation, rather than protected speech in the way xAI claimed.
1
6
Why the judge denied the preliminary injunction
On September 4, Judge Frank denied xAI’s motion. His order did not resolve whether HF 1606 violates the First Amendment. Instead, it concluded that xAI had not demonstrated the immediate, irreparable harm required for the extraordinary remedy of a preliminary injunction.
11
22
Timing mattered. The court had already denied xAI’s temporary-restraining-order request on July 31, noting that the company waited until shortly before the law’s effective date to seek emergency relief. The case then moved on an expedited track for consideration of the preliminary-injunction request.
11
14
In the later ruling, the court also weighed the balance of harms and public interest in Minnesota’s favor. Judge Frank described the statute as a democratically enacted civil measure addressing harms associated with the spread of photorealistic sexualized AI imagery.
5
11
What the ruling means—and what it does not
The practical result is straightforward: Minnesota may enforce HF 1606 while the lawsuit continues. Covered AI providers face the statute’s potential penalties during that period.
5
9
But the decision is not a final endorsement of the law’s constitutionality. The court did not decide xAI’s First Amendment claims on their merits. xAI has said it plans to appeal the denial, while its underlying challenge remains pending.
2
3
Why Grok and “nudification” tools are central to the dispute
The case highlights a broader policy question: whether states can impose direct obligations on AI-product providers when their tools can produce realistic, nonconsensual sexual imagery of identifiable people. Minnesota’s position is that readily accessible nudification features create serious harms and warrant targeted regulation; xAI’s position is that the law imposes a speech-based restriction on general-purpose creative technology.
1
2
17
For now, the procedural ruling favors Minnesota. The more consequential question—whether this type of AI-tool restriction can survive First Amendment scrutiny—remains unresolved.