Internal Turmoil Grows Within Meta's Applied AI Division

Meta is facing significant internal unrest as its newly formed Applied AI team, comprised of approximately 6,500 employees, expresses deep dissatisfaction with their working conditions. Reports indicate that many staff members were forcibly reassigned to this unit, which focuses on generating coding problems and puzzles to train the company's AI models. The environment has become so strained that a recent internal livestream was interrupted by an employee outburst, highlighting the growing frustration among those who feel their roles have been reduced to repetitive, soul-crushing labor.

The friction stems from Meta’s aggressive pivot toward artificial intelligence, which has involved reallocating internal talent rather than relying on external contractors. CEO Mark Zuckerberg has justified this strategy by arguing that Meta’s own engineers possess higher technical proficiency than third-party data labelers. However, the top-down mandate—often delivered via abrupt emails—has left many employees feeling like "draftees" rather than valued contributors, leading to widespread resentment and a decline in company morale.

This situation carries significant implications for Meta’s long-term innovation goals. While the company is pouring billions into AI to maintain its competitive edge, the alienation of its core engineering workforce poses a serious risk to productivity and talent retention. As leadership acknowledges the "distress" caused by these structural changes, the company must now balance its urgent need for high-quality training data with the necessity of maintaining a culture that can attract and sustain top-tier technical talent.