Meta Pauses Employee-Tracking Program After Data Leak
Original: Meta Pauses Employee-Tracking Program Following Internal Data Leak
Why This Matters
Highlights tensions between AI training needs and employee privacy rights in tech; demonstrates internal security vulnerabilities at major platforms.
Meta halted its Model Compatibility Initiative employee-tracking program after an internal security breach exposed sensitive data collected through the system to workers, the company confirmed on June 22, 2026.
Meta paused its divisive Model Compatibility Initiative (MCI) following an internal data exposure incident. The tool, rolled out to US employees in April 2026, collects computer inputs including mouse movements, clicks, keystrokes, and screen content to train AI systems. According to Meta spokesperson Tracy Clayton, "We have carefully designed this program with privacy safeguards and while we have no indication at this time that any data was improperly accessed by Meta employees, we're pausing it while we investigate." On Monday, a Meta engineer issued an internal security notice revealing that MCI databases had been exposed to all company employees. The program initially lacked an opt-out option but changed following worker protests citing privacy and security concerns. Meta executives previously defended the initiative as necessary for training AI to operate computer software like humans, using employees as training examples. Following critical comments in internal forums about the security lapse, Meta announced the pause, informing WIRED before notifying staff. A former employee involved in opposition efforts called the incident "a mess" and criticized leadership for dismissing worker concerns about data safety.