Meta tried to replace workers with AI agents. The agents started breaking the company

The Story

At a Hawaii retreat, Meta hatched Project OT: become “AI native,” let agents do daily work, shrink some teams by up to 60% in two waves. Staff revolted after keystroke/mouse logging was mandated to train replacement agents. Internal posts said unchecked agents took “large-scale, disruptive actions that humans are unlikely to execute.” Major incidents jumped 40%; firefighting time jumped 70%. Code changes +220%, user-facing features only +36%. Zuck canceled the second wave the night before the first layoff.

Why It Matters

The company that wanted to sell AI workers to everyone couldn’t survive its own AI workers. “AI native” as a management religion, then a 40% incident spike. This is future-shock with an org chart.

Evidence

Reuters investigation, 26 Aug 2026, based on internal docs and 20+ sources. Meta confirmed Project OT existed and that some team-cut scenarios hit 60%, calling it “scenario planning.

Sources

Daily scan: 2026-08-27

• reuters.comhttps://www.reuters.com/investigations/mark-zuckerberg-had-bold-plan-replace-meta-staff-with-ai-heres-how-it-imploded-2026-08-26/
• arstechnica.comhttps://arstechnica.com/ai/2026/08/metas-scrapped-plans-to-go-ai-native-included-slashing-teams-by-60-percent/