Image Courtesy: Getty Images
Meta plans to introduce software that monitors employee activity on work devices as part of efforts to improve its artificial intelligence systems. The tool will capture inputs such as mouse movements, clicks, and keystrokes, along with limited screenshots from approved applications.
The initiative was outlined in an internal memo shared with staff, describing the effort as a way to gather real world data on how people interact with computers. The collected data will be used to train AI models designed to perform everyday digital tasks, according to Fortune.
The tracking system will operate only on a defined set of work related apps and websites. Meta says the goal is to help its AI systems better understand practical workflows, such as navigating menus, using shortcuts, and completing routine tasks that current models struggle to replicate.
The project is linked to the company’s broader push to develop AI agents capable of handling white collar work. These systems aim to automate tasks typically performed on computers, but require detailed behavioral data to function effectively.
Meta stated that safeguards are in place to protect sensitive information and that the collected data will be used strictly for AI training purposes. The company emphasized that participation is tied to normal work activity rather than requiring additional input from employees.
The move reflects a wider trend across the technology sector, where companies are seeking new sources of training data to improve AI performance. Reports indicate that other firms, including OpenAI, have explored using real workplace materials to enhance model capabilities.
Meta has significantly increased its investment in artificial intelligence, committing substantial resources to infrastructure and research. The initiative is being developed within its Superintelligence Labs division, which focuses on advancing next generation AI systems.
The approach raises questions about workplace privacy and data use, even as companies argue that such data is necessary to build more capable and practical AI tools. As competition intensifies, access to real world usage data is becoming a key factor in the development of advanced AI systems.
