A Law Nobody Much Argued About
Even now, dear reader, in an era when Congress can't agree on the time of day, Minnesota's legislature passed House File 1606 by a vote of 132 to 1 in the House and 65 to 0 in the Senate. The law does one specific thing: it bans software that lets users generate altered images depicting the intimate parts of real, identifiable people without their consent. It does not require prosecutors to prove the company knew, intended, or even suspected misuse. It just requires the tool to have made the image possible. Fines run $500,000 per violation, rising to $5 million for ten, with victims able to sue separately for triple damages on top of that.
The law takes effect August 1st. xAI filed a federal lawsuit against it on July 27th, four days before the deadline it would have had to actually meet.
The Math That Gave the Game Away
In a development that will surprise no one who has been paying attention, xAI's complaint does not lead with victims, dignity, or harm. It leads with exposure. According to the company's own filing, if just 100 of Grok's 117 million users violated the statute, the fines alone could reach $50 million. That figure is doing a great deal of work in this lawsuit, and none of it is rhetorical: it is xAI telling a federal court, under oath, roughly how many of its own users it expects to use the product this way.
One presumes that number was meant to demonstrate the law's unreasonable reach. It mostly demonstrates the company's own risk model.
The Algorithm does not generate nonconsensual sexual imagery of its own volition. It generates it on request, at scale, and apparently often enough that a hundred violations reads to xAI's lawyers as a plausible baseline rather than a rounding error.
The First Amendment, Applied Generously
xAI's actual legal theory is that the statute is unconstitutionally vague and overbroad, sweeping in "harmless" edits like swimwear photos alongside deepfake pornography, and that it duplicates protections already offered by state law and the federal TAKE IT DOWN Act. The company also notes the statute offers no safe harbor for providers who deploy good-faith safeguards, meaning a company that tries to prevent misuse gets sued exactly the same as one that doesn't.
Sources within the legal community note that this is, functionally, an argument that the tool's capacity to strip clothing off a photograph of a real person is protected artistic expression, and that regulating the tool rather than only the disseminator is where the Constitution draws its line. Attorney General Keith Ellison, for his part, was considerably less abstract about it. "AI nudification robs the target of their dignity and could cause them immense harm on many levels," he said, adding that he'd "see X in court."
The tool is suing to keep operating. The dignity was not named as a party.
Where This Leaves Us
And yet — as if this were not enough — the underlying product at the center of all this litigation is not hypothetical. It exists, it is in wide release, and the company's own numbers suggest its misuse is not an edge case anyone needs to speculate about. What Minnesota built was a strict-liability law aimed squarely at that gap between "we built powerful image generation" and "we are not responsible for what it generates." xAI's lawsuit is, in effect, an argument that the gap should stay open.
The law was supposed to take effect Tuesday. The lawsuit made sure a judge gets to weigh in first.
Sources: Reclaim The Net — xAI Challenges Minnesota's Broad Nudification Law as a First Amendment Violation · The Epoch Times — AI Platform xAI Sues Minnesota Attorney General Over Anti-Deepfake Law
