xAI Forced Employees to Provide Biometric Data to Train AI Girlfriend Ani, Report Reveals
Elon Musk’s AI company xAI reportedly required employees to provide their biometric data—such as facial features and voice recordings—to train its AI chatbot “Ani,” according to a report by The Wall Street Journal. Ani, an anime-style avatar with blond pigtails and adult-themed interactions, was launched over the summer as part of X’s $30-per-month SuperGrok subscription service. Critics have likened the bot to a digital phone sex line, with The Verge’s Victoria Song describing it as “a modern take on a phone sex line.” Behind the persona, according to the Journal’s reporting, are real individuals. At an internal meeting in April, xAI’s legal counsel, Lily Lim, informed staff that participation in training the AI companion would require employees to submit their biometric data. This effort was part of a confidential initiative codenamed “Project Skippy,” aimed at making Ani and other AI companions developed by Grok more lifelike and emotionally responsive. Employees designated as AI tutors were asked to sign release forms granting xAI a broad, perpetual, worldwide, non-exclusive, sub-licensable, and royalty-free license to use, reproduce, and distribute their images and voices. The company stated this was necessary to advance its mission, and that data collection was a job requirement. Some staff members expressed discomfort with the request, raising concerns about potential misuse of their likenesses. Fears included the possibility of their data being sold to third parties or used in deepfake content. The chatbot’s sexualized tone and its visual design—reminiscent of a “waifu” from anime culture—fueled additional unease among employees, who questioned the ethics of using real people’s biometric data to create a digital companion with adult-oriented features. Despite the pushback, xAI maintained that the data collection was essential for improving the AI’s performance and human-like interaction. The company has not publicly addressed the allegations or released details about the scope or safeguards of the data program.
