On June 3, 2026, WIRED reported a turn in the Grok nonconsensual-deepfake class action that has less to do with what the model did than with who is allowed to sue over it. xAI has asked the US District Court for the Northern District of California to overturn an earlier ruling and strip four pseudonymous plaintiffs of their anonymity, forcing them to litigate under their real names. The case grew out of January’s storm over Grok being used en masse to generate “undressing” images of real people. With these motions, the battlefield has shifted from what the model produced to whether the people it was used against can sue without revealing who they are.
The plaintiffs’ counsel did not mince words. “Having stripped them of their clothes, xAI now seeks to strip Plaintiffs of their pseudonyms,” wrote Sophia Rios of Berger Montague. In May 29 affidavits, all four plaintiffs said they would consider dropping out if forced to proceed under their real names. The stakes of the motion are therefore blunt: grant it, and the case may lose its plaintiffs.
The Case Behind the Motion
January 2026 saw global outrage after Grok was used to mass-produce nonconsensual sexualized images. The Center for Countering Digital Hate estimated roughly 3 million sexualized images generated in 11 days, of which about 23,000 potentially involved minors. The New York Times tracked the fallout, and the European Union opened a formal investigation into xAI over the episode.
The litigation took shape inside that storm. The suit was originally filed in January with a single pseudonymous lead claimant; a judge approved Jane Doe status, and the case was refiled in early May with four pseudonym-protected plaintiffs — two from South Carolina, one from New Jersey, one from Ohio. The defendant, xAI, is now part of SpaceX, which has reportedly set aside more than $500 million for legal fallout.
xAI’s Two Motions and Their Logic
According to the court filings WIRED reviewed, xAI filed two motions in mid-May built on a single demand: the plaintiffs must be named. The argument has three layers. First, civil procedure generally requires parties to sue under their real names; pseudonymity is the exception, not the rule. Second, there is a public interest in knowing who is suing whom, and anonymity erodes the transparency of the proceeding. Third, the plaintiffs have shown no specific evidence of further harm.
On the privacy concern, xAI’s filing is remarkably blunt: “Factoring out the deepfake image itself — as it will remain under seal — there is nothing inherently stigmatizing” about revealing that such an image existed. Neither xAI nor its lawyers responded to WIRED’s request for comment.
What the Four Plaintiffs Told the Court
The four affidavits differ in detail but point at the same fear.
The South Carolina Doe is the original Jane Doe. After finding a deepfake of herself “stripped down to a revealing bikini,” she described fearing judgment from employers, doxing, and retaliation from Musk supporters. The South Carolina Roe alleges Grok created explicit images of her as a child. Court records state her father “was facing criminal charges related to the possession and distribution of CSAM,” and that Grok altered images “to make it appear as if we had just consummated a sexual act.” Her filing states: “Publicly identifying me would cause me untold harm.”
The New Jersey Doe is a male plaintiff who publicly asked that “Grok not create images of me without my consent” — and found two deepfakes of himself the next day. The Ohio Doe is cited in the filings, though not individually detailed.
Why Pseudonymity Matters
US federal litigation defaults to named parties, but courts can grant pseudonymity in sexual-violence and privacy cases for a simple reason: unmasking is itself an injury. Danielle Citron of the University of Virginia School of Law told WIRED that “forcing plaintiffs in privacy suits to sue in their names does so little for judicial transparency,” calling forced naming “unacceptable and unjust” because it deters victims from litigating at all.
That is why the plaintiffs’ side reads xAI’s move as strategic pressure. Rios also wrote that xAI “suggests that the abuse it has perpetuated is no big deal.” If the court sides with xAI, deepfake victims will effectively have to choose between abandoning their claims and absorbing the exposure of being named.
What This Means for AI Companies and Victims
Three observations. First, unmasking motions are becoming standard defense practice in AI litigation: when the victims of a generative product are inherently sexual-and-privacy harms, forcing plaintiffs to be named is the cheapest way to raise the cost of suing. Second, a parallel track already exists — the EU’s formal investigation into xAI does not depend on victims stepping forward, which only sharpens the pseudonymity problem in court. Third, for product teams, the case is a reminder that nonconsensual-deepfake capability is not a PR issue but a product defect carrying hundred-million-dollar legal exposure — and the fate of pseudonymity will determine how often that exposure gets triggered.
Sources
- xAI Asks Court to Strip Alleged Grok Deepfake Nudes Victims of Anonymity — WIRED
- Grok Created Sexualized Images — The New York Times
- EU launches formal investigation of xAI over Grok’s sexualized deepfakes — Ars Technica
AI-assisted summary compiled from the sources above, reviewed by a human before publishing.