Meta workers protest mouse-tracking software as AI tensions erupt

Meta employees turned meeting rooms, vending machines and even toilet paper dispensers into protest sites this week, circulating flyers against software that tracked mouse movements, clicks and keystrokes on company computers. The visible leafleting signaled a new level of internal pushback as the Facebook owner presses workers to adjust to an AI-centered operating model.

The flyers, distributed at multiple U.S. offices, urged employees to sign an online petition and used the line, “Don’t want to work at the Employee Data Extraction Factory?” The materials also invoked employees’ rights to organize under the U.S. National Labor Relations Act, framing the dispute as more than a privacy complaint. It was a direct challenge to how much control Meta should have over the data generated on workers’ desks.

Meta said the tracking software was meant to support development of AI agents that can perform work tasks autonomously. Company spokesperson Andy Stone said Meta needs “real examples of how people actually use them” if it is building agents that help users complete everyday tasks on computers. That explanation has done little to quiet concerns that the same data could be used to measure performance, police behavior or accelerate automation inside the company.

The backlash came as Meta was preparing to cut about 8,000 jobs, or roughly 10% of its global workforce, starting May 20, with more layoffs expected later in 2026. That timeline has sharpened employee fears that the company’s AI tools and restructuring plans are moving in lockstep, with the software collecting behavioral data just as management trims headcount.

The protest also pointed to a broader shift across white-collar workplaces in the United States. As companies race to build AI systems that can handle routine digital tasks, workers are increasingly confronting new forms of monitoring that blur the line between product development and surveillance. At Meta, the clash over mouse-tracking software has become a test of whether employees can resist the data collection that increasingly underpins AI management.