Meta Scales Back Employee Keystroke Tracking After Backlash

Meta has revised its controversial plan to monitor employee computer activity after facing significant internal resistance, according to an internal memo sent Tuesday.

The social media giant announced new controls that allow workers to pause data collection for up to 30 minutes at a time and request complete exemptions from the program. The changes come after more than 1,500 employees signed a petition opposing the Model Capability Initiative (MCI), a tool designed to log keystrokes and mouse clicks to train AI models.

The original plan and employee concerns

Meta introduced the MCI in April, explaining that capturing real examples of computer usage would help build AI agents capable of completing everyday tasks. The company emphasized that data would not be used for other purposes and included safeguards to protect sensitive content.

Employees reacted negatively to the announcement. One worker described the tracking as "very dystopian," expressing concerns that the data could be used to justify further job cuts. Meta has eliminated approximately 2,000 positions this year and announced plans in April to reduce its workforce by 10%, affecting roughly 8,000 employees.

Another former employee characterized the tracking tool as "just the latest way they're shoving AI down everyone's throat."

Technical issues compound privacy worries

Beyond privacy concerns, employees reported practical problems with the monitoring software. The tool reportedly consumed substantial computing resources, draining laptop batteries and causing internet usage spikes for remote workers.

Stephane Kasriel, a vice president in Meta's Superintelligence Labs unit, acknowledged these issues in the internal memo. He stated that the team had introduced "several optimizations" to reduce battery impact and responded to employee feedback about personal data on work devices and the desire for greater control over data capture timing.

Why it matters

Meta's retreat on workplace monitoring reflects growing tension between AI development needs and employee privacy expectations. As companies race to train AI models with real-world data, they face pushback from workers who view comprehensive tracking as invasive—especially during periods of layoffs when surveillance feels threatening. The compromise solution of time-limited pauses and opt-outs may set a precedent for how tech companies balance AI training requirements with workforce concerns.

Reuters first reported details of the internal memo and the policy changes.