MENLO PARK, June 23 (Parliament Politics Magazine) – Meta announced on Monday that it is pausing an internal program designed to track employee mouse movements and digital activity for AI training purposes.
The company stated the halt is necessary to investigate emerging data security concerns regarding the initiative.
The social media giant confirmed the pause following revelations that sensitive employee information, originally intended for monitoring digital interactions within internal systems, remained accessible to all company staffers. Internal documents reviewed by Reuters highlighted the extent of the exposure. While Meta acknowledged the ongoing investigation, the company declined to specify the expected duration of the program suspension.
“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,” company spokesperson Tracy Clayton stated.
The initiative, known as the Model Capability Initiative, or MCI, launched in April to capture mouse movements, clicks, and keystrokes on computers used by U.S.-based employees. The collected information was intended to train Meta’s AI models. According to sources familiar with the matter, the tool was still actively recording as of Monday afternoon, though the company noted that the rollout of the pause would take time to implement across the organization.
The decision to stop the program followed a high-priority security incident report filed by a staff member. Documentation indicated that the exposed data included full prompts, transcriptions, private conversations, performance data, and sensitivity ratings. Previous reports in May raised concerns that the program was collecting more information than originally described and storing it in unencrypted formats.
Internal commentary on the security incident report underscored the severity of the situation. One employee noted that they had accessed personal tax and medical information through their work computer, expressing frustration that such sensitive data was not protected by aggressive filtering as previously promised.
Prior to the security breach, the MCI program faced significant criticism from the workforce. Employees reportedly viewed the mandatory surveillance as an intrusive practice, with complaints centering on technical performance issues. Staff members reported that the background software caused excessive battery drain on company laptops and triggered spikes in home internet data usage for those working remotely.
In response to initial complaints, Meta had previously attempted to modify the program by offering employees the option to pause tracking for 30-minute intervals. However, workers indicated that the mechanism was insufficient, leading to continued friction within the company. The program’s reach reportedly extended beyond internal tools, capturing activity when employees visited external sites such as LinkedIn and Google. The current security investigation remains the primary focus as the company addresses the fallout from the exposed data.
