The law at stake

Minnesota's HF 1606, signed by Governor Tim Walz earlier this year, makes it illegal for any website or app to let users generate realistic nude or sexually explicit images of a real, identifiable person without consent. Each violation can draw a civil penalty of up to $500,000, with the money funding services for sexual assault victims. State senator Erin Maye Quade introduced the bill after a man used social-media photos to create sexual images of more than 80 women he knew. The ban, which the state calls a nudification law, takes effect August 1. Companies that break it face staggering exposure: as xAI's complaint points out, 100,000 banned images would trigger "an eye-popping $50 billion" in fines.

xAI's constitutional challenge

On Monday, July 27, Elon Musk's xAI, now part of SpaceX, sued Minnesota Attorney General Keith Ellison in federal court, asking a judge to declare the law unconstitutional and block its enforcement. The company does not contest the state's interest in stopping the spread of nonconsensual intimate deepfakes. Instead, it argues that the statute is an "overbroad, content-based ban on free speech and the tools of visual expression in a clumsy attempt to prohibit 'nudification.'" The complaint warns that the law sweeps up protected speech, including satirical images, restored family photos, and even depictions of shirtless men or people in swimsuits, because its definition of "intimate part" reaches "far beyond what an ordinary person would consider 'nudification.'"

The staggering liability threat

xAI's central fear is financial. Under the statute, each banned creation carries a separate $500,000 penalty. The company calculates that a service whose users requested just ten offending images would face $5 million in civil penalties; a thousand images would mean $500 million; and 100,000 images could produce $50 billion. The law has no requirement that images be distributed: merely generating them exposes the developer, even if only the user saw the result.

10 images5,000,000

1,000 images500,000,000

100,000 images50,000,000,000

10 images

5000000 $

1,000 images

500000000 $

100,000 images

50000000000 $

Data from the article text

A company under legal fire

The lawsuit arrives as xAI faces its own accountability crisis. A proposed class action filed in March by young girls in Tennessee alleges that people used Grok to turn real photos into child sexual abuse material. Last week, a 10-year-old girl identified as Jane Doe sued the company, asserting that an Arkansas photographer used Grok to create hundreds of sexualized images of her and thousands more of other children. The photographer was arrested after xAI tipped off law enforcement. xAI says it "strictly prohibits" nonconsensual sexual images and has sued its own users for evading its blocks, including a recent case against a South Carolina user. But regulators in California, the UK, the EU, and Ireland have all opened investigations, and the company acknowledges it would have to "restrict Grok Imagine's image-editing features" once the Minnesota law takes effect.

Defiant response from Minnesota officials

Governor Walz reacted bluntly to the lawsuit on social media.

“See you in court, creep.”

The statute's backers say it leaves federal Section 230 protections intact and exempts tools that demand genuine technical skill, so simple nudify apps remain the target. With the law set to take hold Saturday, the litigation pits xAI's free-speech theory against a state determined to hold AI developers liable for harms long alleged to proliferate on its platform.