Meta To Record Employee Mouse Movements For AI Training
Meta is preparing to begin collecting employee mouse movements and keystrokes as part of an effort to build training data for its artificial intelligence tools, according to a Reuters report.
The plan involves capturing how employees interact with their work computers, including cursor activity and typed input, in order to generate data that can be used to improve AI systems. The reporting describes the initiative as an internal rollout focused on AI training, not a consumer-facing change.
Meta, the parent company of Facebook, Instagram and WhatsApp, has been investing heavily in AI products and infrastructure. Collecting detailed interaction data from employees would give the company a large, controlled stream of examples of how people complete tasks on a computer, information that can be used to train models intended to assist with similar work.
The development matters because it touches on two high-stakes issues at once: the competitive push to improve AI systems and the boundaries of workplace monitoring. Mouse movements and keystrokes can reveal not only the outcome of work but also the process—how long tasks take, what steps were tried first, and what was corrected along the way. That kind of granular data can be valuable for training AI tools designed to follow instructions, navigate software, or support productivity workflows.
At the same time, collecting that level of detail inside a workplace can raise questions about internal policies, employee notice, and how such data is stored and used. The reports indicate the stated purpose is AI training, placing the initiative within a broader industry effort to gather high-quality data for developing new models and AI assistants.
News organizations including Fortune and Business Insider also reported on the program and described it as tracking activity on employee screens and input behavior. Business Insider reported on a memo related to the rollout, underscoring that the effort is being communicated internally as Meta expands AI development across the company.
What happens next will depend on how the program is implemented across Meta’s workforce and how the collected data is handled within the company’s AI training pipeline. As the initiative rolls out, employee-facing guidance, internal controls, and oversight around data access are likely to be central to how the program operates day to day.
Meta has positioned AI as a core priority, and the reported move to capture employee mouse movements and keystrokes signals how far the company is willing to go to build the training data needed to power its next generation of tools.
