This week, Microsoft has enabled the modern “computer use” function for Copilot Studio, which allows AI agents to interact with websites and stationary applications. Similarly to the identically called “computer” function of the Openai or Claude operator, companies will be able to exploit Copilot Studio to create AI agents that treat websites and computer applications, as if they were tools for automating tasks.
“Using a computer allows agents to interact with websites and computer applications by clicking the buttons by selecting the menu and entering the fields on the screen” Explains Charles LamannaCorporate Vice President of Microsoft Business & Industry Copilot. “This allows agents to support tasks, even if there is no available API interface to directly connect to the system. If a person can use the application, the agent can also.”
Copilot Studio will be able to create AI agents who can automate data entering, carry out market research and even processing invoices. This modern tool will detect even when the buttons and screens change in applications or websites and will continue to work without failure or break, according to Microsoft.
Microsoft added a similar function, named for his consumer Copilot at the beginning of this month. Actions can be carried out in the background while working on other tasks. Copilot can now do things such as booking book restaurants, tickets for events and shopping in online stores. The experience of activities, however, is constrained to many partners and it seems that Copilot Studio will be able to work with more websites and applications.