NEW YORK CITY, NEW YORK: Meta is rolling out internal tracking software that records how employees interact with their computers, as the company seeks to improve its artificial intelligence systems using real-world work behavior.

The program will capture mouse movements, clicks, and keystrokes from U.S.-based employees, according to internal memos seen by Reuters, forming part of a broader effort to develop AI agents capable of carrying out tasks autonomously.

The tool is designed to run across selected work-related applications and websites, while also taking occasional snapshots of employees' screens to provide additional context for training data.

According to one of the memos, shared within Meta's AI-focused Meta SuperIntelligence Labs team, the initiative aims to address gaps in the company's models, particularly in handling everyday computing tasks such as navigating dropdown menus and using keyboard shortcuts.

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

Meta said the data collection would be limited in scope and focused solely on improving its AI systems.

Meta spokesperson Andy Stone said the data collected would not be used for performance assessments or any other purpose other than model training, and that safeguards were in place to protect sensitive content.

"If we're building agents to help people complete everyday tasks using computers, our models need real examples of how people actually use them, things like mouse movements, clicking buttons, and navigating dropdown menus. To help, we're launching an internal tool that will capture these kinds of inputs on certain applications to help us train our models," Stone said.