The Rise of Workplace Surveillance

Meta has begun tracking employee computer activity to feed its generative AI training systems. The tool, called Model Capability Initiative (MCI), logs what workers see and do on their computers—ostensibly to help train next-generation AI models. But the timing, combined with recent waves of layoffs, has sparked internal alarm.

The Privacy Problem

"It feels very dystopian," one Meta employee told the BBC, asking not to be named. The company has already cut around 2,000 workers this year, and employees widely expect deeper cuts ahead. Meanwhile, job listings on Meta's internal career site have plummeted from around 800 openings in March to just seven in April—a stark indicator of where the company is headed.

The difference with MCI is the intent. Meta could always see employee activity on company computers. But tracking and logging that data specifically for AI training is new. And the coincidence—ramping up AI investment while gutting headcount—sends a clear message: the future of Meta's workforce looks automated.

The Numbers Behind the Obsession

CEO Mark Zuckerberg has pledged to spend roughly $140 billion on AI in 2026, nearly double the previous year. The company is positioning itself at the cutting edge of large language models and generative tools, investing heavily in its Meta Superintelligence Labs group and launching models like Muse Spark.

In January, Zuckerberg said 2026 would be "the year that AI dramatically changes the way we work." He suggested projects once requiring large teams could now be handled by a single talented person.

The implicit subtext: Why pay dozens of engineers when AI can do the job?

What Employees Are Saying

Beyond the unnamed employee who called it "very dystopian," another former Meta staffer described MCI as "just the latest way they're shoving AI down everyone's throat." The resentment seems widespread—and justified.

Companies have a right to monitor workplace security and productivity. But there's a line between oversight and harvesting human labor data to train systems that might eventually replace those same workers.

Meta's approach feels like crossing it.

Source: BBC News