Meta’s New AI Push: Employee Keystrokes and Clicks Become Training Data

Meta is introducing a new internal tracking tool that will log how employees use company computers and apps, including keystrokes and mouse clicks, to help train its AI models. The company says the tool is meant to give its systems real examples of how people complete everyday computer tasks, and insists the data will not be used for anything else. Still, the move is causing unease among some workers, especially as it comes alongside layoffs, a hiring freeze, and a broader company-wide push to pour more money and focus into AI.

Key points

Meta told employees that a new tool will run on company computers and internal apps, recording activity to be used as training data for AI technology.

According to Meta, the goal is to improve AI agents by showing them how people actually work on computers in real life.

A company spokesman said the tracking data is not being used for any other purpose and that safeguards exist to protect sensitive content.

Some current and former employees reacted negatively, describing the move as unsettling and as another sign that Meta is aggressively pushing AI into the workplace.

The rollout comes as employees are already worried about more job cuts. The article says Meta has laid off around 2,000 employees this year in smaller rounds, and workers have been expecting deeper losses in the months ahead.

The company also appears to be shrinking hiring. Its jobs site had about 800 listings in March, but now shows just seven. Meta declined to comment on that change or on plans for further cuts.

The tracking tool is called Model Capability Initiative, or MCI, according to Reuters. While Meta could already access employee activity on company devices, using that activity specifically to train and improve AI tools is described as new.

The move fits into Mark Zuckerberg’s larger AI strategy. He has pledged to increase spending on AI, with Meta planning to spend roughly $140bn on AI in 2026, nearly double what it invested a year earlier.

The company also invested $14bn in Scale AI in 2025 and brought in some of its executives to support Meta’s AI development.

Meta’s reworked Meta Superintelligence Labs group released Muse Spark last month, and the company hopes data from the new employee tracking system will help train future models from that lab.

Key quotes

“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 data is not used for any other purpose.”

“But one Meta employee, who asked not to be identified, said having their smallest actions on a computer being used to train AI model as workers expect a slew of additional job cuts feels ‘very dystopian’.”

“This company has become obsessed with AI.”

“Just the latest way they’re shoving AI down everyone’s throat.”

In January, Zuckerberg said 2026 will be “the year that AI dramatically changes the way we work”.

“We’re starting to see projects that used to take big teams now be accomplished by a single, very talented person.”

Implications

This story points to a sharper shift inside Meta: employee work habits are now being treated as valuable raw material for building AI tools. That may help the company develop more capable computer-using AI systems, but it also appears to be deepening worker anxiety at a moment when job cuts and shrinking hiring already have employees on edge.

It also suggests Meta’s AI strategy is not just about outside investments and flashy model launches. The company is now turning inward, using the day-to-day behavior of its own workforce to support the next wave of AI development.

Finally, the article highlights a tension at the heart of Meta’s AI push: the company frames this as necessary innovation, while some employees see it as a troubling sign of how far AI is reshaping work, power, and job security inside the company.

Source: https://www.bbc.com/news/articles/cvglyklz49jo