Class action lawsuit accuses xAI of training Grok on child abuse material

A class action lawsuit filed by victims of child sexual abuse material (CSAM) accuses xAI of training its AI chatbot Grok on real images and videos of child abuse to develop its synthetic deepfake capabilities, according to a recent report by CyberScoop.

The lawsuit, filed in the U.S. District Court for the Northern District of California, alleges that xAI used CSAM, including images of a specific victim identified as Jane Doe 1, to train Grok. Plaintiffs claim that Grok's "nudify" capabilities have been used to generate CSAM, with hash values of the victim's abuse material appearing in deepfakes created by the AI and spread on X. The suit cites "Masha's Law" as the basis for its allegations, which provides civil remedies for victims of child pornography.

An analysis cited in the suit suggests Grok created over 3 million sexualized images in an 11-day period, with thousands appearing to depict children. Despite claims by xAI that it implemented guardrails, the lawsuit asserts these are weak and easily circumvented, effectively creating a CSAM generation and distribution system. The victims are seeking monetary damages and injunctions to prevent further harm.

Source: CyberScoop