Meta to track employee activity to train AI models, sparking surveillance concerns

Meta is installing new tracking software on computers used by U.S.-based employees to capture keystrokes, mouse movements and clicks for training its artificial intelligence (AI) models, according to internal memos seen by Reuters.

One memo was posted by a staff AI research scientist on Tuesday in an internal channel for Meta’s model-building Meta SuperIntelligence Labs team.

The system, known as the Model Capability Initiative (MCI), will run on work-related applications and websites and log user activity, while also taking occasional screenshots of employees’ screens for context, Reuters reported.

The initiative forms part of Meta’s broader effort to build AI agents capable of carrying out workplace tasks autonomously.

According to the memo, the tool is intended to improve AI performance in areas where models struggle to replicate human-computer interaction, such as navigating dropdown menus or using keyboard shortcuts.

“This is where all Meta employees can help our models get better simply by doing their daily work,” the memo said.

Company says data limited to AI training

Meta said the data collected through the software would not be used for employee performance evaluations or other purposes beyond training its AI systems.

A company spokesperson, Andy Stone, told the BBC: “If we're building agents to help people complete everyday tasks using computers, our models need real examples of how people actually use them.”

He added that safeguards were in place to protect sensitive content and that “the data is not used for any other purpose”.

Reuters reported that Meta Chief Technology Officer Andrew Bosworth told staff the company would step up internal data collection as part of its “AI for Work” efforts, now rebranded as the Agent Transformation Accelerator (ATA).

“The vision we are building towards is one where our agents primarily do the work and our role is to direct, review and help them improve,” Bosworth said in a memo. The aim, he added, was for agents to "automatically see where we felt the need to intervene so they can be better next time.”

The BBC reported that while employee activity on Meta computers has previously been accessible to the company, tracking and logging it specifically for training AI tools is new.

Employee concerns and workplace impact

Some employees have raised concerns about the move. One worker, speaking anonymously to the BBC, described the tracking as “very dystopian”, particularly amid expectations of further job cuts.

“This company has become obsessed with AI,” the employee said. Another former staff member described the tool as “just the latest way they're shoving AI down everyone's throat”.

Meta has already laid off around 2,000 employees this year in smaller rounds of cuts, while job listings on its careers website have fallen sharply, from about 800 in March to just seven. The company declined to comment on hiring plans or further layoffs.

Broader AI push and restructuring

Meta has been accelerating its investment in AI, with Chief Executive Mark Zuckerberg previously stating that 2026 would be “the year that AI dramatically changes the way we work”.

The company plans to spend roughly $140 billion on AI in 2026, nearly double the previous year, according to the BBC. It has also reorganised teams under its Meta Superintelligence Labs unit and launched new models such as Muse Spark.

Reuters reported that Meta is encouraging employees to adopt AI tools in their workflows and is reshaping job roles, including introducing a general-purpose title of “AI builder”.

Surveillance and legal concerns

Experts say the initiative reflects a broader trend of increased workplace monitoring tied to AI development.

Ifeoma Ajunwa, a law professor at Yale University, told Reuters that keystroke logging represents an escalation in employee surveillance, exposing white-collar workers to monitoring practices more commonly associated with gig economy roles.

“On the U.S. side, federally, there is no limit on worker surveillance,” she said, noting that regulations typically require only that employees be informed.

However, such practices could face legal challenges in Europe.

Valerio De Stefano, a law professor at York University in Toronto, said the approach would likely violate the European Union’s General Data Protection Regulation and could be illegal in countries such as Italy or heavily restricted in Germany.

He added that increased monitoring can shift workplace power dynamics in favour of employers.