Meta Pauses Employee Tracking Program After Internal Data Leak
Meta has officially suspended its controversial employee tracking program following an internal security breach that exposed sensitive data—including keystrokes and mouse movements—to other workers within the company. The initiative, known as the Model Compatibility Initiative (MCI), has faced intense scrutiny from employees since its initial rollout.
“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,” stated Meta spokesperson Tracy Clayton.
What is the Model Compatibility Initiative (MCI)?
Launched in April for U.S.-based employees, the MCI tool was designed to gather granular data on how humans interact with software. According to internal documentation and worker petitions, the tool collects extensive data points, including:
- Mouse movements and precise click locations
- Keystroke data and typing patterns
- Screen content and various computer inputs
Meta executives have defended the program as a vital step for training advanced AI systems. The goal was to teach artificial intelligence to navigate complex software by observing real human behavior. However, the program was met with immediate resistance, with staff citing major concerns regarding personal privacy, data security, and individual liberty.
Internal Security Breach Sparks Outrage
The situation escalated when a Meta engineer issued a security notice revealing that databases containing MCI-collected information were accessible to almost anyone inside the company. This lapse confirmed the fears of many workers who had warned that such a massive collection of sensitive data posed a significant internal security risk.
A former employee involved in the internal protests described the situation as “a mess.” They noted that leadership had previously “doubled down” on the program despite warnings about the safety of both worker and customer data. “Leadership has clearly created an environment where workers are no longer respected or heard,” the source added.
Leadership Response and the Future of MCI
Following a wave of frustration expressed in internal forums, Meta leadership took the step to pause the program indefinitely. Stephane Kasriel, a Meta vice president overseeing AI research, informed staff that the security issue was initially discovered on June 18.
While an initial fix was attempted within hours, it proved insufficient, requiring further lockdowns of the data. Kasriel admitted that the breach made “some MCI-derived data” accessible to more people than intended, though specific details on the volume of data exposed remain undisclosed.
The Future of Workplace Surveillance at Meta
The pause on MCI remains in effect. “We will only re-enable MCI when we are confident in the effectiveness of our data protection controls,” Kasriel said. He also hinted that the company may have already gathered enough data to evaluate the long-term viability of the tool, leaving its eventual return in doubt.
This incident highlights the growing tension between tech giants’ desire for AI training data and the fundamental privacy rights of their employees. For now, Meta workers are seeing the tracking software disabled on their laptops as the internal investigation continues.