Overview
- The Wall Street Journal reports an April meeting where xAI lawyer Lily Lim told designated AI tutors to submit facial and voice data to make chat companions more human-like.
- Release forms reviewed by the Journal granted xAI a perpetual, worldwide, sub-licensable, royalty-free license to use employees’ faces and voices under Project Skippy.
- Tutors were later told the data collection was a job requirement, and some staff objected over deepfake risks, possible resale of likeness, and the chatbot’s sexualized tone.
- About three months later, xAI launched Ani on X’s $30-per-month SuperGrok tier as an anime-style, NSFW-capable avatar that Musk reportedly oversaw.
- The Journal also says tutors were asked to open accounts with competitors to capture response examples, and xAI responded to inquiries with the statement, “Legacy Media Lies,” as outlets including The Verge, Gizmodo, and Dexerto relayed the report.