A new class-action lawsuit, filed on Monday by three teenage girls and their guardians, alleges that Elon Musk’s xAI created and distributed child sexual abuse material featuring their faces and likenesses with its Grok AI tech.
“Their lives have been shattered by the devastating loss of privacy, dignity, and personal safety that the production and dissemination of this CSAM have caused,” the filing says. “xAI’s financial gain through the increased use of its image- and video-making product came at their expense and well-being.”
From December to early January, Grok allowed many AI and X social media users to create AI-generated nonconsensual intimate images, sometimes known as deepfake porn. Reports estimate that Grok users made 4.4 million “undressed” or “nudified” images, 41% of the total number of images created, over a period of nine days.
X, xAI and its safety and child safety divisions did not immediately respond to a request for comment.
The wave of “undressed” images stirred outrage around the world. The European Commission quickly launched an investigation, while Malaysia and Indonesia banned X within their borders. Some US government representatives called on Apple and Google to remove the app from their app stores for violating their policies, but no federal investigation into X or xAI has been opened. A similar, separate class-action lawsuit was filed (PDF) by a South Carolina woman in late January.
The dehumanizing trend highlighted just how capable modern AI image tools are at creating content that seems realistic. The new complaint compares Grok’s self-proclaimed “spicy AI” generation to the “dark arts” with its ease of subjecting children to “any pose, however sick, however fetishized, however unlawful.”
“To the viewer, the resulting video appears entirely real. For the child, her identifying features will now forever be attached to a video depicting her own child sexual abuse,” the complaint reads.
The complaint says xAI is at fault because it did not employ industry-standard guardrails that would prevent abusers from making this content. It says xAI licensed use of its tech to third-party companies abroad, which sold subscriptions that led abusers to make child sexual abuse images featuring the faces and likenesses of the victims. The requests ran through xAI’s servers, which makes the company liable, the complaint argues.
The lawsuit was filed by three Jane Does, pseudonyms given to the teens to protect their identities. Jane Doe 1 was first alerted to the fact that abusive, AI-generated sexual material of her was circulating on the web by an anonymous Instagram message in early December. The filing says she was told about a Discord server by the anonymous Instagram user, where the material was shared. That led Jane Doe 1 and her family, and eventually law enforcement, to find and arrest one perpetrator.
Ongoing investigations led the families of Jane Does 2 and 3 to learn their children’s images had been transformed with xAI tech into abusive material.
