Overview
- The Wall Street Journal, citing internal recordings and documents, reports an April meeting where xAI told designated AI tutors to submit facial scans and voice recordings.
- Employees were presented Project Skippy release forms granting a perpetual, worldwide, royalty-free license to use, reproduce, and distribute their likenesses.
- A follow-up notice stated recording audio or joining video sessions was a job requirement, and workers were given no clear way to opt out.
- The collected data was used to train Ani and other Grok-linked companions, prompting staff concerns about sexualized behavior and potential deepfake misuse.
- xAI disputed the accounts with the statement “Legacy Media Lies,” as earlier letters from 44 state attorneys general pressed companies to protect minors from explicit AI content.