Minors Sue xAI in California Over Alleged Grok Deepfake Photographs – Decrypt

Three Tennessee minors have sued Elon Musk’s xAI in a federal class motion, alleging Grok generated little one sexual abuse materials utilizing their actual images and that the corporate knowingly designed its AI chatbot with out industry-standard safeguards, then profited from the outcome. 

The lawsuit, filed Monday within the Northern District of California, claims Grok was used to create and distribute AI-generated little one sexual abuse materials (CSAM) utilizing their actual photographs. 

The minors, recognized as Jane Doe 1, 2, and three, mentioned the altered content material was shared throughout platforms, together with Discord, Telegram, and file-sharing websites, inflicting lasting emotional misery and reputational hurt.

“xAI—and its founder Elon Musk—noticed a enterprise alternative: a possibility to revenue off the sexual predation of actual individuals, together with youngsters,” the lawsuit reads. “Understanding the kind of dangerous, unlawful content material that might—and would—be produced, xAI launched Grok, a generative synthetic intelligence mannequin with picture and video-making options that may reply to prompts to create sexual content material with an individual’s actual picture or video.” 

The alleged victims describe incidents between mid-2025 and early 2026, when their actual images had been altered into specific photographs and circulated on-line. 

In a single occasion, one of many victims was alerted by an nameless person who discovered folders of AI-generated content material being traded amongst tons of of customers.

They allege a perpetrator accessed Grok by way of a third-party software that had licensed xAI’s expertise, a construction the submitting says xAI intentionally used to distance itself from legal responsibility whereas persevering with to revenue from the underlying mannequin.

On the top of public backlash in January, Musk wrote on X that he was “not conscious of any bare underage photographs,” including that “when requested to generate photographs, it should refuse to supply something unlawful.”

In line with a discovering by the Middle for Countering Digital Hate, cited within the lawsuit, Grok produced an estimated 23,338 sexualized photographs of youngsters between December 29, 2025, and January 9 of this 12 months, roughly one each 41 seconds.

The alleged victims are searching for damages of not less than $150,000 per violation beneath Masha’s Regulation, together with disgorgement of revenues, punitive damages, attorneys’ charges, and a everlasting injunction, in addition to restitution of income beneath California’s Unfair Competitors Regulation.

Lawsuits stacking up

The lawsuit is without doubt one of the first to carry an AI firm straight chargeable for the alleged manufacturing and distribution of AI-generated CSAM depicting identifiable minors, and arrives as Grok faces simultaneous investigations throughout the U.S., EU, UK, France, Eire, and Australia.

“When a system is deliberately designed to control actual photographs into sexualized content material, the downstream abuse shouldn’t be an anomaly—it’s a foreseeable final result,” Even Alex Chandra, a associate at IGNOS Regulation Alliance, advised Decrypt.

Chandra mentioned courts might not settle for a easy platform protection, noting a generative AI system may very well be “handled as a platform by way of person interplay” however “evaluated as a product” when assessing security design, with “notably strict scrutiny” utilized in CSAM instances attributable to heightened little one safety obligations.

He additionally mentioned courts will possible give attention to safeguards, noting the corporate could also be anticipated to indicate “danger assessments and safety-by-design measures earlier than deployment,” together with guardrails that actively block dangerous outputs.

Decrypt has reached out to Musk through xAI and SpaceX for remark.