Mark Zuckerberg's plan to cut up to 60% of staff through AI agents was scaled back dramatically after internal pushback and underwhelming results
Mark Zuckerberg wanted to rebuild Meta from the inside out, replacing large swaths of human workers with AI agents and creating what he called an “AI native” organization.
Project OT, hatched during a leadership retreat in Hawaii in January 2026, envisioned workforce reductions of up to 60% across two phases in May and November. Instead, Meta ended up cutting roughly 10% of its workforce and publicly admitting that AI technology wasn’t ready for the role it had been cast in.
The plan versus what actually happened #
The concept behind Project OT was ambitious: AI agents would take over daily tasks previously handled by humans, with smaller “talent-dense” teams overseeing the bots.
Meta executed a 10% workforce reduction around May 20, 2026. A second, larger wave was scheduled for November. That wave never came.
Strong employee resistance halted the November layoffs at the last moment. The company pivoted to a far more modest reduction the following day. Around 6,500 employees were transferred into a newly created Applied AI unit, reoriented toward AI-focused roles rather than being shown the door.
The numbers told a complicated story about what AI agents were actually delivering. Code changes surged 220% year-on-year. But user-facing feature updates rose by just 36%. The increased code volume came with an alarming rise in technical issues.
Morale cratered as surveillance concerns mounted #
Employee sentiment dropped from 74% to 55% during the Project OT rollout. Workers flagged specific concerns about surveillance and job security tied to tracking software deployed alongside the AI agents.
The tracking tools were designed to monitor productivity and identify roles suitable for AI replacement.
Zuckerberg’s public admission #
By July 2026, Zuckerberg publicly acknowledged that the pace of AI advancements was slower than anticipated. The admission led Meta to pull back from the most radical elements of Project OT while continuing to invest heavily in AI infrastructure and training.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our