This week, Microsoft has rolled out a new feature to its Copilot Studio that allows artificial intelligence agents to interact with websites and desktop applications. Similar to the “computer use” capabilities found in platforms like OpenAI’s Operator and Claude, this feature enables businesses to develop AI agents that can utilize websites and applications for task automation.
Charles Lamanna, corporate vice president of Microsoft’s business and industry Copilot, described the functionality in a recent announcement. He stated, “Computer use enables agents to interact with websites and desktop apps by clicking buttons, selecting menus, and typing into fields on the screen. This allows agents to handle tasks even when there is no API available to connect to the system directly. If a person can use the app, the agent can too.”
With the capabilities of Copilot Studio, AI agents can efficiently manage tasks such as data entry, conducting market research, or processing invoices. The tool is also designed to adapt to changes in apps and websites, automatically adjusting to new buttons and interfaces without interruption, according to Microsoft.
Earlier this month, Microsoft introduced a related feature called Actions to its consumer Copilot product. This feature allows tasks to be executed in the background while users continue their work on other projects. For instance, Copilot can now handle restaurant bookings, event ticket purchases, and online shopping. However, the Actions functionality is limited to specific partners, while Copilot Studio appears to be more versatile in its compatibility with various applications and websites.
