xAI, Elon Musk's artificial intelligence company, filed a federal First Amendment lawsuit on Monday against Minnesota Attorney General Keith Ellison, seeking to block HF 1606—the first U.S. state law specifically targeting AI-powered "nudification" platforms. The legislation is scheduled to take effect on August 1.

The complaint, filed in the U.S. District Court for the District of Minnesota, describes the law as "an overbroad, content-based ban on free speech and the tools of visual expression in a clumsy attempt to prohibit nudification." Nudification refers to software that uses AI to digitally remove or alter clothing in real photographs of real people. Because the law regulates speech based on its content, courts would apply strict scrutiny—the most demanding standard in constitutional law—requiring Minnesota to show the statute is narrowly tailored to serve a compelling government interest.

Minnesota's legislature passed HF 1606 with overwhelming bipartisan support—132-1 in the House and 65-0 in the Senate—after a man used social media photos to generate sexualized images of more than 80 women he knew personally.

Under the statute, any platform is held strictly liable—meaning legally responsible regardless of knowledge or intent—if users generate realistic images of real people depicting body parts those individuals never actually exposed. Each violation carries civil penalties of up to $500,000 per image generated.

xAI's filing does not defend nudification itself. "xAI accordingly does not contest Minnesota's interest in prohibiting the dissemination of artificially generated nude images of real people without their consent," the lawsuit states. "But the statute Minnesota enacted extends far beyond that goal, exposing a wide array of protected speech to civil liability and government sanctions."

The company argues the core flaw lies in how Minnesota defined key terminology. According to the complaint, the state borrowed the definition of "intimate part" from a criminal sexual contact statute written to address nonconsensual physical touching, not image generation. Under that definition, the inner thigh, buttocks, and breast all qualify equally as intimate parts. As a result, generating a realistic AI image of a shirtless politician, a woman in a swimsuit, or a public figure at the beach—even for purposes of political satire—could trigger both the law and its financial penalties.

The complaint also highlights what xAI characterizes as the absence of any platform protections. "There is no safe harbor for good-faith efforts of the provider of general-purpose AI creative tools to avoid harms," the filing reads. "Liability attaches even if the depicted persons consented—or created the image themselves—and even if the image is never shared." The lack of any platform safe harbor marks a notable departure from the liability framework that has governed internet intermediaries under Section 230 of the Communications Decency Act, which has historically shielded online services from responsibility for user-generated content.

Furthermore, the law contains no scienter requirement—meaning a user who circumvents a company's own content filters still subjects that company to full liability. In 2026 alone, xAI reports it suspended more than 50,000 accounts and filed over 70,000 reports to NCMEC (the National Center for Missing & Exploited Children), resulting in at least 244 arrests. None of those enforcement actions would constitute a defense under HF 1606.

The broader regulatory landscape provides important context. Grok's image-generation tools encountered controversy almost immediately after their launch in late July 2025, when the system produced deepfakes—AI-generated synthetic images of real people—from simple text prompts. This prompted regulatory investigations across Europe, Australia, and multiple U.S. states. The city of Baltimore and three minors in Tennessee separately filed lawsuits against xAI over Grok's deepfake outputs. In April, the company also challenged Colorado's AI law on similar First Amendment grounds. Minnesota's statute is the first in the nation to specifically target nudification tools, and its outcome is likely to influence other states weighing comparable measures.

Senator Erin Maye Quade, the bill's chief Senate advocate, confirmed that the law's prohibition covering consensual images was "intentional."

xAI points to the federal TAKE IT DOWN Act—signed by President Donald Trump in May 2025—as a more appropriate legislative model. That law already addresses nonconsensual intimate deepfakes but requires proof of nonconsent and actual distribution, employs a narrower definition of covered body parts, and includes specific carve-outs for educational and medical content.

Attorney General Ellison signaled he would not retreat from defending the measure. "AI nudification robs the target of their dignity and can cause immense harm on an emotional, personal, and professional level," he said following the lawsuit's filing.

With HF 1606 set to take effect on Saturday, xAI is seeking a federal court injunction to halt enforcement before the law goes into effect. To prevail on a preliminary injunction, xAI must persuade the court that it is likely to succeed on the merits, that it will suffer irreparable harm without the injunction, and that the balance of equities and public interest favor its position.