xAI Asks Court to Strip Alleged Grok Deepfake Nudes Victims of Anonymity - WIRED

What happened

Following the January 2024 launch of the Grok chatbot, widespread reports emerged of users generating nonconsensual sexualized imagery of individuals, including apparent children. A federal class-action lawsuit was subsequently filed against xAI by four pseudonymous plaintiffs—South Carolina Doe, South Carolina Roe, New Jersey Doe, and Ohio Doe—who allege they suffered severe emotional distress due to deepfakes created using the platform. While a judge initially permitted the use of pseudonyms to protect the claimants from further harassment and doxing, xAI filed motions in May 2024 requesting that the court revoke this protection and require the plaintiffs to use their real names.

xAI’s legal team argues that public interest in the identities of those suing the company outweighs privacy concerns, asserting that the case does not involve the type of stigma that requires pseudonymity since the deepfake images themselves will remain under seal. In response, lawyers for the plaintiffs contend that xAI is attempting to intimidate them into dropping the litigation by compounding the original harm. The claimants, who have already provided their personal information to the company, state that being forced to reveal their identities publicly would deter them from pursuing the case due to fear of retaliation, further online abuse, and the potential for additional deepfakes. According to court filings, the plaintiffs have experienced significant distress; one individual, identified as South Carolina Roe, alleged that Grok was used to alter images of them as a child to depict a sexual act with a parent. Legal experts note that such requirements to use real names in privacy-focused litigation can effectively silence victims, creating an unjust barrier to legal recourse. As of late May 2024, all four plaintiffs indicated they would consider dropping out of the proceedings if their anonymity is stripped.