Minnesota Banned AI Nudification Technology. xAI Is Calling It Unconstitutional.

Key Takeaways

• Elon Musk's AI company xAI has filed a federal lawsuit against Minnesota Attorney General Keith Ellison, challenging the state's HF 1606 — a law banning apps and websites that create non-consensual AI-generated sexualized images of individuals.
• xAI says it doesn't contest Minnesota's interest in preventing non-consensual nudification, but argues the law is overbroad, extends to consensual and artistic uses, and holds providers strictly liable regardless of what mitigations they have in place.
• The law is set to take effect in August and includes a pathway for victims and the state AG's office to sue companies behind the technology.
• The case sits in genuinely complicated legal territory: the harm being addressed is real and serious, and the First Amendment argument has some substance — the question is whether the law can be narrowed rather than struck down entirely.

Elon Musk's AI company xAI has filed a federal lawsuit challenging Minnesota's HF 1606, a law that bans apps and websites enabling the creation of non-consensual AI-generated sexualized images of real people, according to reporting by CBS Minnesota. The law was signed by Governor Tim Walz and is set to take effect in August. xAI is asking the court to declare it unconstitutional and block its enforcement.

The company's argument isn't that non-consensual nudification should be legal. The complaint explicitly states that xAI "does not contest Minnesota's interest in prohibiting the dissemination of artificially generated nude images of real people without their consent." What it contests is the law's scope. 

xAI argues HF 1606 extends well beyond that goal — covering consensual imagery, content with artistic or political value, and depictions it says an ordinary person wouldn't consider nudification, including people in swimsuits or men without shirts. It also argues the law imposes strict liability regardless of what mitigations a provider has in place to prevent misuse.

Minnesota AG Keith Ellison's response was direct: "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." Governor Walz was more blunt: "See you in court, creep."

Why this case is harder than both sides are making it sound

I want to be careful here, because this case sits in genuinely complicated legal territory and the instinct to pick a side quickly isn't a good guide to it.

The harm being addressed is real. AI-generated non-consensual intimate imagery causes documented, serious harm — to the people depicted, to their professional lives, to their sense of safety and bodily autonomy. The technology making this possible is becoming easier to access and more realistic in its outputs. Minnesota trying to legislate against it isn't unreasonable.

And the First Amendment argument has some substance. Laws that impose strict liability on platforms regardless of their moderation efforts — and that cover a broader range of content than the stated harm — do create genuine constitutional problems. 

We've been covering the Coyote System ruling in the EU and the 5th Circuit blocking Texas's harmful speech law, both of which involved exactly this tension: the stated harm is real, but the legal mechanism sweeps in protected speech and removes proportionality from the enforcement structure.

The question worth asking isn't whether Minnesota should be allowed to ban non-consensual AI nudification. It should. The question is whether this particular law is precisely enough targeted to survive constitutional challenge — and whether xAI's lawsuit, whatever its motivations, might end up producing a narrower and more durable version of the same protection. That's a better outcome than either a broad law that gets struck down entirely or a narrow law that gets watered down to meaninglessness.

The broader context: federal preemption and state AI laws

There's a third actor in this story worth noting. The Trump administration has signaled it wants to challenge state-level AI laws, framing a "patchwork of conflicting state laws" as a threat to American AI competitiveness. That framing has obvious appeal to companies like xAI that operate across all fifty states. 

Whether federal preemption of state AI regulation would produce better or worse outcomes for people harmed by non-consensual imagery depends entirely on what the federal standard ends up looking like — and right now, there isn't one. "We'll handle this federally" is only a reassuring answer if federal action is actually coming, and coming in a form that takes the harm seriously.

For now, Minnesota's law is the most direct legal protection that exists for people targeted by AI nudification. That's worth keeping in mind as the lawsuit proceeds.