Meta Platforms is rolling out new tracking software on US-based employee computers that captures mouse movements, clicks and keystrokes for artificial intelligence training, tying everyday workplace behavior directly to the company’s effort to build AI agents that can complete work on computers.
According to Reuters, the tool runs only on a list of work-related apps and websites, and also takes occasional screen snapshots for context. Meta spokesperson Andy Stone said the collected data is intended only for model training, not employee performance assessments, and said safeguards are in place to protect sensitive content.
Internally, the company is framing the program as a way to solve a specific weakness in current AI systems: navigating software the way real workers do. The memo cited by Reuters said the goal is to improve models on tasks such as using dropdown menus and keyboard shortcuts, with the company effectively sourcing real-world human-computer interaction data from its own workforce.
Sponsored · Selkirk Copper Mines
“If we’re building agents to help people complete everyday tasks using computers, our models need real examples of how people actually use them,” Stone said, adding that the internal tool will capture inputs on certain applications to train those models.
The timing fits a broader Meta spending cycle that is increasingly dominated by AI infrastructure. Meta reported $200.97 billion in 2025 revenue, up 22% year over year, while capital expenditures surged to $72.22 billion for the year. For 2026, the company guided to an even larger $115 billion to $135 billion in capital expenditures, with management saying the increase will support Meta Superintelligence Labs efforts and the core business.
Meta also ended 2025 with $81.59 billion in cash, cash equivalents and marketable securities, generated $115.80 billion in operating cash flow for the year, and said headcount rose 6% year over year to 78,865 as of December 31, 2025. At the same time, Meta is reportedly preparing layoffs beginning May 20 that are expected to affect about 10% of its workforce, or roughly 8,000 employees, as it restructures more aggressively around AI.
Information for this story was found via the sources and companies mentioned. The author has no securities or affiliations related to the organizations discussed. Not a recommendation to buy or sell. Always do additional research and consult a professional before purchasing a security. The author holds no licenses.
