Meta's Workforce Replacement Plan: AI Agents Caused Chaos
Disruptive effects of AI agents forced Meta to change its plans.
1Meta's Radical Transformation Plan
Earlier this year, Meta explored scenarios to reduce some teams by up to 60% under its 'AI native' initiative called Project OT. The plan envisioned two waves of layoffs. Internal documents obtained by Reuters revealed AI was meant to take over daily tasks of thousands of employees, with small human teams overseeing the AI systems.
2First Wave of Layoffs and Reversal
The first layoff wave occurred in May, but CEO Mark Zuckerberg canceled the November wave. Reuters, citing over 20 sources, couldn't determine exactly why Zuckerberg changed course. Meta confirmed it shelved the project before finalizing how many positions would be affected.
3Disruptive Effects of AI Agents
Internal communications showed AI agents took 'large-scale, disruptive actions unlikely to be executed by humans,' leading to a 40% increase in major technical and security incidents compared to the previous year. Employee time spent resolving these issues increased by up to 70%. Meta declined to comment on these internal posts.
4Productivity Gains Fell Short
According to CTO Andrew Bosworth's early June post, code changes to internal platforms increased 220% year-over-year. However, new or improved features reaching users only grew by 36%. These metrics indicated AI failed to deliver expected productivity improvements.
5Impact on Employee Morale
Layoff rumors in March-April and monitoring of keyboard/mouse inputs for AI training (later paused) negatively impacted morale. Meta specifically emphasized that performance ratings and promotions were determined by people, not AI.
6AI Native Vision and Third-Party Sales
Project OT documents defined 'AI native' as an environment where 'AI-ready tools interact, workflows are automated, and new builds are AI-first.' In June, Meta began selling these AI agents to third parties as part of this vision.
7Pilot Program and Middle Management Reduction
Meta launched a pilot program restructuring at least 10 teams including engineering and research. An October 2023 internal 'AI-Native Playbook' post outlined plans to reduce middle management and use 'agent-assisted analysis' for daily prioritization.
8Industry Implications
Meta's experience demonstrates even the most eager organizations may struggle to replace human workloads with AI. The case highlights risks of overambitious AI projects while serving as a cautionary tale for the industry. AI's potential to transform workforce dynamics appears not yet mature enough for radical implementations.