Working inside the tech giant is looking less and less like the safe Silicon Valley utopia you might remember from the past. According to global media reports, Meta’s US employees have organized grassroots protests against their employer’s latest data harvesting methods. The core of the conflict is software logging employee activity on an unprecedented scale.
The corporation has rolled out an internal tool called the Model Capability Initiative (MCI), part of the broader Agent Transformation Accelerator program (formerly known as “AI for Work”). This system is designed to constantly log mouse movements, clicks, and keystrokes, and even take periodic screenshots of employee computers.
Company management is officially defending the move, arguing that this kind of behavioral data is essential for building the next generation of digital assistants. According to this narrative, for algorithms to smoothly navigate software and understand complex processes, they must learn directly from the behavioral patterns of living experts. A Meta representative tried to calm the waters in a media statement:
“If we’re building agents to help people complete everyday tasks using computers, our models need real examples of how people actually use them”.
The corporate spokesperson also assured that the gathered information will absolutely not be used for performance reviews, and the system itself has built-in safeguards to protect sensitive content.
These technological arguments, however, are completely failing to win over the staff. Employee frustration stems directly from the broader business context. Meta is currently on the brink of another drastic restructuring – according to industry reports, the company plans to lay off nearly 8,000 people (roughly 10% of its workforce) to offset massive spending on AI infrastructure.
Protesters are therefore raising a powerful ethical argument. They point out that traditional workplace monitoring is used to manage a workforce that ultimately keeps its jobs. In this case, experts are being forced to hand over their institutional knowledge and problem-solving frameworks developed over years to machines that are ultimately meant to make them redundant.
Agreeing to be hired, according to the employees, does not equate to agreeing to serve as commercial training fodder from which the corporation sucks out competencies right before handing out pink slips.