Meta Platforms has internally announced a controversial new data-collection program that will monitor every mouse click, keystroke, and screen movement of its U.S. employees.
The initiative, titled the Model Capability Initiative (MCI), represents a radical shift in how the tech giant sources training data to build autonomous AI workplace agents.
Meta Recording the “Human Touch” in Computing
According to internal memos, Meta’s current AI models struggle with nuanced human-computer interactions, such as navigating complex dropdown menus or utilizing specific keyboard shortcuts.
To bridge this gap, MCI will log granular behavioral data to create high-fidelity datasets.
The goal is to train AI agents that can seamlessly mimic human workflows, effectively performing administrative and technical tasks with human-like precision.
Meta CTO Andrew Bosworth further detailed a re-branded “Agent Transformation Accelerator” (ATA) program.
He noted that the company aims for a “closed-loop” system where AI agents perform the bulk of the work, while human employees shift into roles as directors and reviewers.
Read Also: Wipro Hires 7,500 Freshers in FY26 but Skips Next Year’s Targets
Surveillance Concerns and Safeguards
The announcement has sparked intense debate regarding workplace privacy.
While Meta spokesperson Andy Stone clarified that the data is strictly for model training and will not influence performance reviews, employees remain skeptical.
Legal experts suggest that logging biometric input dynamics—such as the trajectory of a cursor or typing cadence—occupies a murky legal territory compared to standard productivity monitoring.
The timing of the rollout has added to internal anxieties.
Meta is reportedly preparing for significant global layoffs in May 2026, leading some staffers to fear they are effectively being asked to “train their own replacements.”
A Massive Financial Bet on AI
This data-harvesting push aligns with Meta’s aggressive 2026 financial roadmap, which includes a projected $140 billion investment in AI infrastructure.
By utilizing its own workforce as a living laboratory, Meta hopes to bypass the “quality ceilings” of synthetic data and the dwindling supply of public internet information.
Note: We are also on WhatsApp, LinkedIn, and YouTube to get the latest news updates. Subscribe to our Channels. WhatsApp– Click Here, YouTube – Click Here, and LinkedIn– Click Here.
About the Author
Sahiba Sharma
Contributing Writer