Meta announced on Monday that it will temporarily suspend an internal initiative that logs employee mouse movements and other digital activity for the purpose of training its artificial intelligence systems. The move comes as the company examines concerns about the protection and exposure of employee information tied to the project.
Meta described the program - known internally as the Model Capability Initiative, or MCI - as a tool deployed in April to capture mouse movements, clicks and keystrokes on U.S.-based employees' computers to help improve its AI models. The company stated it had implemented privacy safeguards when designing the program and that, at present, it has no indication that Meta staffers improperly accessed the data. Nevertheless, Meta said it will pause the initiative while it conducts an investigation.
"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, were pausing it while we investigate," said company spokesperson Tracy Clayton.
The decision to halt MCI followed a high-priority security incident report, known internally as an SEV, filed by an employee who raised concerns about the exposure of employee information. Internal documentation tied to the SEV lists data that was exposed, including "full prompts and transcriptions, private conversations, people & performance data, DSS sensitivity ratings (1-4)."
Documentation related to the SEV also records an employee request for deeper scrutiny. One employee wrote that they had accessed personal tax and medical information through a work computer and noted that many thousands of colleagues had similar access. The employee said they had been told such data would be protected and used only for valid business purposes following aggressive filtering.
Separately, reporting in May indicated the MCI program collected more information than initially described and retained some of that information unencrypted, which raised privacy concerns among staff. According to internal materials, the tool was still recording as of Monday afternoon, though Meta said the pause was rolling out and would take time to reach all affected systems.
Meta did not specify how long the suspension would remain in effect. The company emphasized it was undertaking an internal review to determine whether safeguards operated as intended and whether any additional steps were required to protect employee data.
Context and implications
The pause affects an internal AI training pipeline that relies on direct capture of user interactions on company devices. The SEV and internal comments underscore employee concerns about whether sensitive personal information was sufficiently protected and about visibility into how internal monitoring tools are configured and audited. Meta's statement frames the action as precautionary while the company seeks to verify whether any internal access was improper.