A fourteen-page memorandum opinion and order signed by Donovan W. Frank, United States District Judge for the District of Minnesota, entered September 4, 2026 as Doc. No. 54 in X.AI LLC v. Ellison, Civil No. 26-3425 (DWF/DTS) (Case 0:26-cv-03425-DWF-DTS). It denies xAI's motion for a preliminary injunction against Minn. Stat. § 325E.91, the statute enacted as H.F. 1606, which prohibits an owner or controller of a website, application, software, program or other service from allowing a user to nudify an image or video or from nudifying one on a user's behalf. The case is tracked at X.AI LLC v. Ellison (challenging Minnesota's nudification-technology statute).
Disposition
The motion is denied on two independent grounds — unreasonable delay coupled with a failure to show irreparable harm, and the balance of harms and public interest — and the court expressly declines to reach the merits of the First Amendment claim. The order states that xAI's motion for preliminary injunction (Doc. No. 11) is "respectfully DENIED."
The statute as the court describes it
The opinion quotes subdivisions 2 and 3 of § 325E.91 in relevant part: a person who owns or controls a service must not "allow a user to access, download, or use" it "to nudify an image or video" or "nudify an image or video on behalf of a user," with an exemption where the service "requires the technical skill of a user to nudify an image or video." Subdivision 1(d) defines nudification as a process by which an image or video "is altered or generated to depict an intimate part not depicted in an original unaltered image or video of an identifiable individual" and where the result "is so realistic that a reasonable person would believe that the intimate part belongs to the identifiable individual." "Intimate part" takes the meaning given in Minn. Stat. § 609.341, subd. 5 — "the primary genital area, groin, inner thigh, buttocks, or breast of a human being." The attorney general may enforce the section and seek civil penalties up to $500,000 for each violation together with injunctive and other relief, under § 325E.91, subd. 5 and Minn. Stat. § 8.31, subd. 3.
The opinion records the enactment sequence by citation to the legislative journals: the Minnesota House passed the bill on April 23, 2026 (Minn. H.J., 94th Leg., Reg. Sess. 6139-40 (2026)); the Senate passed it unanimously on April 29, 2026 (Minn. S.J., 94th Leg., Reg. Sess. 9257-58 (2026)); the Governor signed it on May 7, 2026 (Minn. H.J. 7315); and it took effect August 1, 2026.
The legislative record recited
Several pages summarize the evidence before the Minnesota legislature, material the court later relies on in the public-interest analysis.
At a February 19, 2026 Commerce Finance and Policy Committee hearing, three women testified to harms following a perpetrator's use of nudification technology on their likenesses — missed work, fear for family members and children, fear of leaving the house, and "the emotional toll of wondering who has seen or who will see convincing, sexually graphic images and videos of them." That perpetrator created non-consensual pornographic images and videos of over 80 women. RAINN told the committee that in one month alone in 2023 there were 24 million users of nudification apps, and that in 2023 over 95 percent of manipulated images were nonconsensual, sexually explicit, and featured women. The Internet Watch Foundation identified a "pedophile guide" encouraging use of nudification tools to generate material and to blackmail children, and reported 6,800 reports flagging child pornography created with generative AI in the first half of 2024, rising to over 400,000 reports through June of 2025.
At a March 26, 2026 Judiciary Finance and Civil Law Committee hearing, RAINN's Senior Legislative Policy Counsel submitted a study of the first 11 days after a social media site launched a one-click AI image-editing tool: approximately 65 percent of the photos created were photorealistic sexualized images, amounting to 3 million such images at a pace of 190 images per minute, of which roughly 23,000 were of children.
xAI and Grok Imagine as the record presents them
Drawing on the complaint (Doc. No. 1) and a declaration (Doc. No. 14), the opinion describes Grok Imagine as an image and video generation model in the Grok family, available in all fifty states, that returns an image or video in seconds from a typed or dictated description. xAI's pleaded position is that outputs span "art, animated family photographs, memes, product mockups, political satire, and religious imagery," that its terms of service prohibit "any illegal, harmful, or abusive activities," and that its policy bars nudification of a real person's image, altering a real person's image to depict them in an intimate or sexual context, and sexualizing or exploiting children. A footnote records xAI's claim that in 2026 alone it suspended over 50,000 accounts and made over 70,000 reports to the National Center for Missing & Exploited Children, resulting in at least 244 arrests.
Against that, the opinion records the State's evidence that these measures are insufficient, citing three suits alleging that photographs of children were used to create child pornography on Grok Imagine — Doe 1 v. X.AI Corp., No. 5:26-cv-2246 (N.D. Cal., amended complaint July 7, 2026); Does v. X.AI, No. 4:26-cv-750 (E.D. Ark., amended complaint August 17, 2026); and Does v. X.AI, No. 4:26-cv-772 (E.D. Ark., amended complaint August 17, 2026) — treated together at Grok CSAM suits (Doe plaintiffs v. X / xAI). It also notes that SpaceX, which the opinion describes as xAI's parent company, "has set aside a fund with more than $500 million for potential litigation losses stemming from Grok Imagine's 'Spicy' and 'Unhinged' modes," citing Maxwell Zeff and Paresh Dave, "SpaceX Listed Grok's 'Spicy' Mode as a Risk in Its IPO Filing," Wired (May 20, 2026).
Delay and irreparable harm
The court applies the four-factor standard of Winter v. Natural Resources Defense Council, 555 U.S. 7 (2008) and Dataphase Systems v. C L Systems, 640 F.2d 109 (8th Cir. 1981) (en banc), noting that the movant bears the "complete burden" (Gelco Corp. v. Coniston Partners, 811 F.2d 414, 418 (8th Cir. 1987)) and that "a failure to demonstrate irreparable harm, standing alone, may be a sufficient basis to deny preliminary injunctive relief" (Caballo Coal Co. v. Indiana Michigan Power Co., 305 F.3d 796, 800 (8th Cir. 2002)).
On diligence, the opinion observes that the statute was signed May 7, 2026 and that xAI moved "[n]early three months later and only three days before the law was set to take effect," offering no explanation "other than to generally suggest that it takes time for laws like H.F. 1606 to reach the attention of corporate management and for management to green light the resources to challenge the law." Acknowledging "no bright-line rule marking when a motion for a preliminary injunction must be filed," the court finds "no serious dispute that xAI is a sophisticated and well-resourced litigant with the means to move quickly when its interests are at stake," and that the failure to do so "strongly suggests a lack of irreparable harm." A footnote records that the same delay grounded the July 31, 2026 denial of a temporary restraining order (Doc. No. 21).
Three asserted harms are rejected in turn:
- The penalty. The $500,000-per-violation exposure is "monetary in nature" and could be recouped if the statute were later invalidated; the court adds that assessment is "unlikely because xAI has disabled its nudification tool in Minnesota" (Doc. No. 40 ¶ 19).
- Commercial injury. Product counsel team lead Naser Baseer attested that on August 1, 2026 xAI implemented additional controls on Grok Imagine's image-editing feature for Minnesota users, having designed, tested and deployed geographically targeted controls within the image-editing pipeline, and that between July 24 and August 1, 2026 engineers spent "a substantial amount of their time" on that work. The court finds the adjustment "already been made" with no evidence of future cost, the resource claim "vague and conclusory," and the projected loss of Minnesota users from a moderation rate risen "over 10 percent" to be "purely speculative," resting on "a single declaration." Because the statute "applies to all companies providing its users with AI-tools that allow for nudification," the court reasons, "there will be no loss of users vis-à-vis any other competing company."
- First Amendment harm per se. Citing Maixner v. Eidsness, 182 F.4th 736 (8th Cir. 2026), the court holds that the Elrod v. Burns rule applies where First Amendment interests "are either threatened or in fact being impaired at the time relief was sought," but that a movant must still make a clear showing of likely irreparable harm, quoting Benisek v. Lamone, 585 U.S. 155 (2018): in a First Amendment case "a preliminary injunction does not follow as a matter of course from a plaintiff's showing of a likelihood of success on the merits." It notes that a court may independently deny relief for unreasonable delay even in the First Amendment context, citing Hotchkiss v. Cedar Rapids Community School District, 115 F.4th 889 (8th Cir. 2024) and Kohls v. Ellison, 166 F.4th 728 (8th Cir. 2026).
Balance of harms and public interest
Because the injunction is sought against the government, the two factors merge (Morehouse Enterprises v. ATF, 78 F.4th 1011 (8th Cir. 2023)). The court finds the balance "tips steeply in favor of the State," which has "enacted, democratically and nearly unanimously, a civil statute aimed at protecting the public from the undisputed harm stemming from AI-nudification technology and the resulting proliferation of photorealistic sexualized images." It characterizes the legislative choice to reach the tool rather than the user as "a direct response to the unique harms stemming from the new technology and the State's conclusion that curtailing user violations after harmful images are circulated is not enough," and quotes Maryland v. King, 567 U.S. 1301, 1303 (2012), for the proposition that a State enjoined from effectuating its statutes "suffers a form of irreparable injury."
Merits reserved
The conclusion records that "[t]he parties sharply contest the strength of xAI's First Amendment claim," with the State "challenging xAI's standing to bring a First Amendment claim on behalf of its users, disputing xAI's own First Amendment interest at stake, and arguing that H.F. 1606 is valid under any level of scrutiny." The court writes that "[t]he constitutional issues raised by the parties are complex, particularly when considered in the context of this new technology and the risks that it poses to the public," that they "deserve, and will receive, full consideration," and that a determination "will be made in the near future because the State has filed a motion to dismiss that places these issues front and center." It adds that it "may also consider whether xAI is entitled to permanent injunctive relief on the merits."
The opinion does not mention the Department of Justice statement of interest of August 17, 2026 that Minnesota HF 1606 (Nudification Technology Prohibition) records from Politico.
Provenance
Retrieved from a Thomson Reuters legal-documents mirror at fingfx.thomsonreuters.com, rather than from PACER or CourtListener, because no RECAP copy of Doc. No. 54 was located. The document carries its own ECF header, page stamps ("CASE 0:26-cv-03425-DWF-DTS Doc. 54 Filed 09/04/26 Page N of 14") and the judge's signature block. Courthouse News Service hosts an identical fourteen-page copy at courthousenews.com. Reuters reported the same disposition, judge and quoted language on September 4, 2026 (Source: reuters.com), and Minnesota Attorney General Keith Ellison announced the denial publicly the same day.
Relationships
- supports: X.AI LLC v. Ellison (challenging Minnesota's nudification-technology statute)
- supports: Minnesota HF 1606 (Nudification Technology Prohibition)
- depends-on: Minnesota HF 1606 — Prohibition on Nudification Technology (bill text, 2nd engrossment)
- related: Grok CSAM suits (Doe plaintiffs v. X / xAI)
- related: xAI
- related: Grok (xAI)
- related: Algorithmic Speech Doctrine
- related: Synthetic Media / Deepfakes
- related: AI and Children
- related: State-Level AI Regulation