Meta Deploys Employee Monitoring Tool to Capture Data for AI Training
Key Takeaways
- ▸Meta's Model Capability Initiative (MCI) captures granular employee behavioral data—mouse movements, keystrokes, clicks, and screenshots—for AI training
- ▸The program is designed to develop autonomous AI agents that can perform work tasks independently, a key focus of Meta's AI capability development
- ▸The deployment raises critical concerns about workplace privacy, employee surveillance, data ownership, and the absence of apparent consent mechanisms
Summary
Meta is deploying new tracking software called the Model Capability Initiative (MCI) on U.S. employee computers to capture mouse movements, clicks, keystrokes, and periodic screenshots of screen content. The data collection is intended to train AI models, specifically for building autonomous AI agents capable of performing work tasks independently, according to internal memos shared with Reuters by Meta staff.
The tool will monitor activity on work-related applications and websites, operating as part of Meta's broader SuperIntelligence Labs initiative to develop more capable AI systems. According to the reporting, employees were notified of the program through internal channels in late April 2026.
The initiative has drawn significant criticism for its invasive surveillance practices and raises fundamental questions about employee privacy, workplace autonomy, and the ethics of using personal behavioral data to train AI systems. The announcement comes as Meta prepares for a 10% global workforce reduction beginning May 20, adding to concerns among employees about the company's practices.
- The initiative affects U.S.-based employees and is monitored across work-related apps and websites
