Meta has suspended a mandatory AI training program following an internal data leak that exposed sensitive employee information company-wide, according to a Business Insider report on June 22. The incident has ignited fresh tension between the company and its workforce over privacy and trust.

The Data Leak

The program, called the Model Capability Initiative, was introduced in April and designed to collect data on employees' work activities to improve Meta's AI systems. Data gathered included keystrokes, mouse movements, transcripts of conversations and performance-related information. The company reportedly made participation mandatory for most staff.

Screenshots obtained by Business Insider revealed that the collected data was more broadly accessible within Meta than intended. The exposed information included private employee conversations, performance metrics and activity records. Meta classified the incident as a SEV 2 on its internal severity scale where SEV 0 is the most critical.

Employee Backlash

Employees had already expressed discomfort with the idea of their keystrokes and mouse movements being recorded for AI training. Now they face the reality that their data may not have been properly protected. The information was widely accessible across the company instead of being restricted to intended viewers.

Screenshots reviewed by Business Insider showed employees criticizing the failure to secure the data from the start. "I am incensed," one employee wrote in an internal group according to the report. Another employee said there was no evidence of malicious access but called the lack of promised restrictions "super frustrating."

Why This Matters

The incident highlights a growing tension at Meta between the company's aggressive push into AI development and its workforce's privacy expectations. Meta has cut thousands of jobs in part to fund the massive AI infrastructure behind more powerful systems. Building these models requires vast amounts of training data and Meta turned to its own employees to supply it, a move most employees reportedly opposed.

Now employees have learned that the data they were compelled to hand over was not adequately secured. The program remains paused while Meta investigates. A company spokesperson told Business Insider that no evidence of improper access by Meta employees has been found but the company is pausing the initiative anyway.

This is not the first privacy-related controversy for Meta. The company has faced multiple regulatory fines and investigations over data handling practices across its platforms. The internal leak suggests those same issues may extend to how the company treats its own staff.

  • Program Scope: The Model Capability Initiative tracked keystrokes, mouse movements, conversations and performance data for AI training.
  • Data Exposure: Sensitive employee information became broadly accessible company-wide due to inadequate access restrictions.
  • Internal Fallout: Employees expressed anger and frustration over privacy violations, with many opposing the mandatory program from its start.