Meta’s newly formed Applied AI unit, comprising approximately 6,500 engineers and product managers, is in open rebellion, with employees comparing their forced reassignment to a “gulag.” The unrest escalated this week when an anonymous worker hijacked an internal livestream, launching an expletive-laden attack on a senior AI executive.
According to reports from Wired, staff forcibly moved into the unit — with the ultimatum to join or quit — describe generating coding puzzles for AI training as “soul-crushing.” In an internal memo on Friday, CEO Mark Zuckerberg admitted the reorganization had “caused distress” and acknowledged mistakes, promising no further company-wide layoffs in 2026 after a year that saw a 10% workforce reduction and thousands shifted into AI roles.
Zuckerberg detailed plans to restore stability, including scaling back manager-to-employee ratios that had reached 50-to-1, allowing transfers back to original teams where possible, and organizing a company-wide hackathon. Chief Product Officer Chris Cox called recent months “brutal” and urged leaders to “get in touch with the company again.”
The discontent extends beyond the Applied AI unit, with more than 1,600 employees signing a petition against keystroke and click monitoring used for AI training. The turmoil highlights a growing friction in the tech industry, where the AI arms race is straining workforces and raising questions about talent retention and innovation.
