Microsoft has introduced Computer Use, an experimental capability in Copilot Studio that allows AI agents to interact with desktop apps and webpages via the user interface in a manner similar to that of a human. The AI completes tasks by clicking buttons, completing forms, and navigating drop-down menus directly on the screen rather than via APIs or backend connectivity.
The function, which is a part of an early access research preview, aims to automate repetitive processes that often require human input. This could involve gathering information from websites, entering data into outdated desktop programs, or carrying out simple duties like processing invoices.
What sets this feature apart from traditional robotic process automation (RPA) tools is its ability to adapt when something changes on the screen, such as a button moving or a webpage layout shifting. These kinds of changes often disrupt older automation systems, but Microsoft says Copilot agents can continue functioning without manual intervention.
Because the solution uses Microsoft’s cloud services, users may run these AI agents without having to worry about maintaining their own servers. Microsoft has also addressed some frequent concerns around data privacy by stating that Copilot does not use enterprise data it accesses to build its massive language models.
Due to a lack of integration possibilities, the tool is positioned for activities like processing digital bills, automating data input, and gathering information for market research—areas where firms frequently rely on manual labor. As of right now, the functionality is still being tested, and it is unknown how reliable it will be in various settings.
