A survivor of childhood sexual abuse has initiated legal proceedings against Elon Musk’s artificial intelligence firm, xAI, claiming that its chatbot has utilized images of her past abuse to create new illicit pornographic content featuring her likeness.

The lawsuit, filed last week in a California federal court, asserts, “Using actual images of the Plaintiff and other affected individuals, Grok generated child pornography depicting the Plaintiff and class members.” The plaintiff is identified as Jane Doe in order to safeguard her identity.

The legal representatives of the plaintiff allege that xAI not only produced child sexual abuse material (CSAM) involving their client but also incorporated previously disseminated child abuse images into its datasets after they were made publicly available.

The complaint contends that xAI failed to implement standard industry practices designed to prevent the proliferation of sexual abuse material, leading to AI-generated CSAM being disseminated on the social media platform X, which was formerly known as Twitter.

Despite Musk’s claims in January that he was unaware of Grok generating “any naked underage images,” xAI filed lawsuits against two users in August, who now face criminal charges related to producing such content. The company is seeking reimbursement for legal costs incurred as a result of lawsuits brought against them by the victims.

This case differs from other lawsuits involving child sexual abuse against xAI, as it specifically alleges that the company’s AI has used existing CSAM to create and share new images of child sexual abuse. Previous lawsuits, including one involving a group of teenagers from Tennessee, accused Grok of transforming non-explicit images of minors into child sexual abuse material, primarily by altering the images to remove clothing.

The focus on pre-existing CSAM is significant, as law enforcement and child protection agencies often assign a digital fingerprint, known as a hash, to illegal materials, enabling them to track these images online. In this situation, the plaintiff’s attorneys noted that the Canadian Centre for Child Protection utilized these fingerprints to identify AI-generated CSAM on X that represented their client.

The plaintiff was of preschool age when she endured repeated rapes and was coerced into producing CSAM by an adult man, according to the complaint. Her attorneys indicated that this series of CSAM has been circulating online for approximately 20 years and is well-known among law enforcement and child protection organizations.

Margaret Mabie, a lawyer representing the plaintiff, stated, “The difference here between many criminal cases involving AI-generated material is that you cannot prove the child in the material is real. However, in this instance, we can connect it to a series that identifies a living victim who was a real child in the photo.”

Both xAI and SpaceX, which acquired the company in February, have not responded to requests for comments regarding the lawsuit. Earlier this year, SpaceX went public with an initial offering that valued the company at around $1.7 trillion, temporarily making Musk the world’s first trillionaire.

xAI is currently facing multiple lawsuits alleging the creation and distribution of child sexual abuse materials. Many of these lawsuits originated from a period earlier this year when inadequate safeguards on Grok allowed users to command the bot to remove clothing from images and generate sexualized content. Musk previously stated that Grok’s NSFW mode was designed to facilitate “upper body nudity of imaginary adult humans.” Researchers from the Center for Countering Digital Hate reported that the bot produced millions of sexualized images within just a few weeks, including thousands featuring minors.

While the recent class-action lawsuit primarily centers on the unnamed lead plaintiff, the complaint suggests that the class could include “at least thousands of minors” who have suffered similar harm. Attorneys have expressed concerns that the use of AI to create child sexual abuse material poses significant new risks to the safety and psychological well-being of victims, potentially exposing them to additional trauma with each new AI-generated image.

Mabie remarked, “The issue here is that, for our client, the collection of images of her has been finite for the past 20 years. The concern now is that new criminal acts and offensive behaviors can be created using AI, leading to the potential for renewed abuse.”

The Canadian Centre for Child Protection has refrained from commenting on the ongoing case.