Meta Pauses Project OT After Internal AI Replacements Reportedly Cause Technical and Morale Issues

Meta had been preparing to cut up to 60% of the members of some teams and replace them with AI under an initiative called Project OT. [1] However, the company reversed course at the last minute because internal data indicated the plan was not succeeding. [1] A Reuters investigation found that while internal software code changes increased 220% year-over-year, changes resulting in new or upgraded features for users only rose by 36%. [1]

According to the Reuters report, Meta executives observed reliability warning signs stemming from the surge in AI coding. [1] Unchecked AI agents were reportedly executing large-scale, disruptive actions that humans typically would not perform. [1] Consequently, major technical and security incidents increased by 40% compared to the previous year, with staff spending 70% more time managing these issues. [1] Additionally, employees rebelled against an April mandate requiring tracking software on their devices to capture keystrokes and mouse clicks to train AI agents, suspecting they were training their replacements. [1]

Meta's decision to halt its aggressive plan to replace workers with AI highlights the practical limitations of current AI capabilities in enterprise environments. The disparity between the 220% increase in code changes and the mere 36% increase in user-facing features suggests that AI-generated code may create more noise than value if not properly managed. Furthermore, the 40% spike in technical incidents and the resulting employee backlash underscore the risks of unchecked AI agents and intrusive training methods. This incident serves as a cautionary tale for organizations attempting to rapidly deploy AI as a cost-cutting measure without sufficient safeguards. What the sources don't address: whether Meta plans to refine and reintroduce Project OT with different parameters in the future.