Minnesota has effectively banned AI pornography.
The state’s new law prohibiting access to AI nudification technology came into effect on August 1st. This law targets the publishers of websites and software whose AI technology is used to alter an image such that it now displays an “intimate part” (which is defined in section 609.341, subdivision 5 as “includ[ing] the primary genital area, groin, inner thigh, buttocks, or breast of a human being”). While applauded by many as the first law of its kind, Minnesota’s new legislation also has opponents. The first major trial is a legal battle with Elon Musk’s xAI.
Minnesota law bans pornography created using AI
Chapter 72 of Minnesota Session Laws was introduced as HF 1606 by Representative Jessica Hanson (DFL-Burnsville) in the spring of this year. It reads:
A person who owns or controls a website, application, software, program, or other service must not (1) allow a user to access, download, or use [aforementioned software] to nudify an image or video; or (2) nudify an image on behalf of a user.
Software providers that are found guilty of allowing users to nudify an image will face a civil penalty up to $500,000, besides compensatory payments to the victim of an amount up to three times the damages sustained. Importantly, this law does not apply to software that requires technical skill, but only artificial intelligence programs. There are already Minnesota and federal laws prohibiting the technological creation of nonconsensual sexual images, but this new law is unique in its specific target against AI companies.
xAI files lawsuit against Minnesota Attorney General in reaction
On July 27th, xAI filed a lawsuit against Keith Ellison, in his role as Attorney General of Minnesota, to combat the new legislation. In the brief, the company argues that this law ignores private company jurisdiction and violates the first amendment:
xAI strictly prohibits its users from generating nude or sexualized images of people without their consent and has indeed filed suit against users who evade its extensive technological blockers to generate such images in violation of this strict prohibition…the statute Minnesota enacted extends far beyond that goal, exposing a wide array of protected speech to civil liability and government sanctions.
While xAI respects the sentiment behind this new Minnesota law, the company holds that its vague wording threatens to inhibit otherwise harmless free speech. Their lawsuit indicates three major flaws with the new law:
(1) The law does not account for AI providers’ own anti-nudification policies (xAI’s acceptable use policy prohibits “[u]ndressing or nudifying real persons, or otherwise altering a real person’s image or likeness to depict them in an intimate or sexual content”).
(2) The law punishes AI providers for images which their users create consensually and do not disseminate.
(3) The law includes an overly broad definition of “intimate parts” which xAI argues will apply to “ordinary depictions of men without shirts, people in shorts or swimsuits, and other body parts routinely displayed in public.”
Importantly, Chapter 72 singles out the provider of the service for punishment, rather than the individual who actually generates the image displaying an intimate part. A company could theoretically be penalized if a user were to nudify an image of themselves. As xAI details in the lawsuit, they have already begun taking measures in their software’s programming to protect themselves from this law. For example, Grok will not generate an image of a private individual wearing a speedo or similar limited garments. The company has indeed changed its software in Minnesota; while images of public figures like Donald Trump in a speedo can still be generated, Grok refuses to generate a similar image of a private citizen. According to the company, these self-made restrictions clash with their users’ right to free speech.
xAI’s lawsuit does not question the fundamental legality of the generation of pornographic materials by artificial means. Instead, the company focuses on their fear that clearly non-pornographic images (i.e. a woman in shorts, a man in a speedo) will be labeled as having been nudified and thus will be blocked. It is the broad condemnation of AI generated images that conflicts with the first amendment, they argue.
Protecting individuals from nonconsensual AI nudification
There is strong legal precedent for the illegality of the creation and dissemination of nonconsensual pornographic images. The question is, are private regulations enough to protect individuals from exploitation, or must the state step in?
xAI claims to have a strict user policy prohibiting the creation of nonconsensual and child sexual material. Their lawsuit claims that, in 2026, company has suspended over 50,000 accounts and made more than 70,000 reports to the NCMEC (National Center for Missing and Exploited Children), leading to at least 244 arrests. xAI has also sued a user for interfering with the software in order to bypass restrictions and generate child sexual abuse material. They claim that Minnesota’s new AI law is superfluous.
While the company uses this data to prove that they are working hard to combat sexual abuse, the fact that xAI has had to suspend and report so many accounts demonstrates the failure of their safeguard system. This suggests that more legislation is necessary. In 2025, xAI introduced Grok “Spicy Mode” which easily created nude deepfakes of celebrities. It is not the only one; a simple Google search reveals countless similar AI programs designed specifically to nudify images. In 2026, Grok “Spicy Mode” features have largely been locked down, but its creation is extremely questionable. These details raise questions as to xAI’s determination to protect individuals, especially women and children, from predators.
Lawsuit likely to succeed
Nevertheless, legal precedent in related cases makes xAI likely to win this lawsuit against the Attorney General of Minnesota. This law’s extremely broad definition of “intimate part” penalizes items that fall under protected speech. Furthermore, Minnesota already has legislation, such as statute 617.262, prohibiting the creation and dissemination of “deep fakes.” That law’s definition of “intimate part” is much more specific to sexual matters than the one found in new law. There is also a federal law with a more specific definition of its target that was devised to combat nonconsensual nudification of images. The TAKE IT DOWN Act prohibits dissemination of nonconsensual intimate photos, whether authentic or edited, and requires sites to remove them within 48 hours of the complaint. This act is enforced by the Federal Trade Commission. These prominent examples of similar legislation call into scrutiny the wording of Minnesota’s new law. Furthermore per legal precedent, if a government body aims to restrict some kind of speech, it must do so with minimal interference. This is known as the “least restrictive test.” Chapter 72’s broad definition of “intimate part” suggests a level of restriction that is higher than Minnesota lawmakers can justify.
Finally, 47 U.S. Code § 230 gives service providers immunity from culpability for criminal uses of their technology, particularly if the provider has complied with federal regulations and placed adequate restrictions on inappropriate uses of their technology. xAI points out that Chapter 72 makes no reference to such restrictive efforts made by the provider.
The issue of consent
Chapter 72 is unique among these examples in that it explicitly punishes the provider of the service rather than the user, and it penalizes these providers for consensual and nonconsensual images alike. xAI argues that this conflation does not account for religious, scientific, and other otherwise unharmful personal uses of their technology.
However, AI technology is different from other programs in that it is constantly learning from itself. Therefore, even if a nude image was generated or uploaded with the consent of those involved, this image could feed the algorithm and aid in the production of a nonconsensual image later on.
The enduring debate about the legality of pornography
During the research process for this article, I typed “AI used to create nudes” into a search engine in order to find relevant news articles. Instead, all results on the first page had titles such as “best AI nude generators” and “AI remove clothes from a photo instant results.” This is telling.
The prevalence of pornography in modern culture seems to indicate a conflation between liberty and license. Does freedom make any private action good? Addictions to pornography are linked to depression, violence, and decreased satisfaction in real relationships among many other harmful effects.
Pornography, even if consensually created, rests on dubious constitutional grounds. Years of court cases have led to confusing legal precedent. Obscenity is not protected under the first amendment, and the Miller Test is the current tool used to determine whether or not something is considered obscene. It must (1) appeal to prurient interest (2) be patently offensive and (3) lack serious literary, artistic, political, or scientific value. Depending on how it is interpreted, pornography may not pass the Miller Test.
When the founders wrote the first amendment to the Constitution, I have a hard time believing that they would allow it to justify the creation and dissemination of nude images and sexual content. The Miller Test and basic morality agree: AI pornography should be illegal in the United States. The pornography industry facilitates the exploitation of children and damages the minds of those addicted to its product.
Moving forward
If xAI’s lawsuit fails and Minnesota’s new law continues unchanged, AI nude generators will not be able to operate in Minnesota. This will likely raise a new conversation about the constitutionality of pornography in general, as the law does not make a distinction between consensual and nonconsensual images. While one might argue that this is a win for morality and criminal justice, this new law’s broad definition of “intimate part” could be manipulatively interpreted and lead to the effective banning protected speech.
Regardless, the truth is that women and children are being exploited through the use of AI nudification software. Individuals must be held accountable for their actions, and providers must be held accountable for their technology. Minnesota’s new law might be the incentive companies need to strengthen their own safeguards.