AI
Meta Pauses Employee-Tracking Program After Internal Data Exposure
The company suspended its Model Compatibility Initiative following a security breach that exposed keystroke and screen-capture data to internal workers.
WIRED — AI
Meta has paused its Model Compatibility Initiative (MCI), a controversial employee-monitoring program, after an internal security lapse exposed collected worker data across the company.
The tool, rolled out in April to US employees, gathered keystroke logs, mouse movements, click locations, and screen content to train AI systems. While the company initially made the program mandatory, employee backlash led to an opt-out option—though a limited one.
On Monday, a Meta engineer flagged that MCI's databases had become accessible to any company employee, prompting the pause. The company said there was "no indication at this time that any data was improperly accessed," but executives decided to suspend operations while investigating the breach.
What employees are saying#
A former Meta worker deeply involved in opposing MCI called the security failure "a mess" that staffers had predicted. "When workers raised concerns, leadership doubled down and failed to acknowledge the risks," the person told WIRED, characterizing the company culture as increasingly "authoritarian."
Meta executives had defended the data collection as essential for training AI systems to mimic human computer use, arguing employees were the ideal training dataset. But the security incident—which sparked critical backlash in internal forums—prompted the company to announce the pause to WIRED before formally notifying staff.
Read more: WIRED's full coverage

