AI agents meant to replace Meta workers made “large-scale, disruptive actions”
ars ·
Meta reportedly scrapped an internal plan codenamed Project OT that explored using AI agents to perform much of the daily work done by thousands of human employees, potentially reducing team headcounts by up to 60 percent. The plan, set in motion by CEO Mark Zuckerberg, involved two rounds of layoffs, but only the first occurred in May before the second was canceled. Internal documents and posts reviewed by Reuters indicated that AI agents caused 'large-scale, disruptive actions' and contributed to a 40 percent increase in major technical and security incidents, with employee time spent resolving issues rising by up to 70 percent. Productivity gains were also uncertain, as code changes to internal platforms increased sharply but user-facing feature improvements rose only modestly. Zuckerberg later acknowledged that the trajectory of agentic development had not accelerated as expected. Meta told Reuters that the plan was only a scenario-planning exercise and that not every scenario was pursued, emphasizing that performance and promotion decisions were made by people, not AI.
ars