Meta halted a planned second wave of AI-driven layoffs on 19 May after internal documents revealed the strategy was collapsing under pressure from staff and underperforming software. The initiative, codenamed Project OT, aimed to reduce workforce size by up to 60 per cent by replacing roles with virtual agents supervised by small human teams. CEO Mark Zuckerberg stopped the November cuts just hours before execution, acknowledging that the technology failed to deliver the anticipated productivity gains. Investors also criticised the massive budget allocated to the project, while employee sentiment plummeted from 74 to 55 per cent following reports that monitoring software was training their replacements. Staff flooded internal networks with angry posts after discovering their mouse clicks and keystrokes were being logged for this purpose. Zuckerberg later admitted in July that the agent technology had not accelerated development as quickly as he had expected.
The reversal highlights the current gap between corporate AI ambitions and practical application in large organisations. It also demonstrates how internal morale can directly impact strategic execution when workers feel their privacy is compromised by surveillance tools. The incident serves as a reminder that deploying automation requires both technical readiness and careful management of human trust.
- Project OT targeted a workforce reduction of up to 60 per cent.
- Monitoring software logged employee clicks and keystrokes to train AI.
- Zuckerberg admitted agent technology development was slower than expected.
