Prime Highlights
- Meta Platforms has started a program in the United States to collect real computer usage data from employees to improve its AI systems.
- The company aims to build better tools that can understand and handle everyday computer tasks more effectively.
Key Facts
- Meta is tracking actions like mouse movements, clicks and typing while employees work on selected applications and websites.
- The collected data will be used to improve AI models and is not intended for employee performance evaluation.
Background
Meta Platforms has started a new internal program in the United States to improve its artificial intelligence systems. It is collecting real computer use data from employees.
The company is tracking actions like mouse moves, clicks and typing. This data will help train its AI models to understand better how people do everyday digital work.
The company stated in internal communication that the purpose of this step is to improve areas where its AI systems still face challenges, such as selecting options from dropdown menus and using keyboard shortcuts. It added that the data will help build AI agents that can perform everyday computer-based tasks more efficiently.
Meta is using a tool that runs on selected work applications and websites. The system collects interaction data while employees carry out normal work activities. The company clarified that this data is aimed at improving model performance and not for employee evaluation.
A company spokesperson explained that if AI agents are being built to help users complete tasks on computers, they need real examples of how people interact with systems. This includes actions like clicking buttons, using menus and pressing shortcuts and the data will help build better and more useful tools.
The move reflects Meta’s broader focus on developing advanced AI systems that can assist users in real-world work environments. The company aims to improve productivity tools and build systems that can handle tasks on their own more easily and correctly.
