Microsoft has unveiled a new feature for its Copilot Studio, called “computer use,” which allows AI agents to autonomously access websites and applications. This innovation is designed to help companies build AI agents without the need for coding expertise.
Through Copilot Studio, users can simply input their desired actions in natural language, test, and fine-tune the instructions in a simulated environment to see the agent in action. The feature enables AI agents to automate tasks on user interfaces across various desktop and browser applications, such as Edge, Chrome, and Firefox.
Charles Lamanna, CVP of Business and Industry Copilot, explained in a blog post that with the computer use feature, enterprises can build agents that interact with applications by performing actions like clicking buttons, selecting menus, and typing into fields. This capability allows agents to connect with systems that don’t have APIs, offering greater flexibility in automation. Furthermore, Microsoft hosts the infrastructure, so organizations don’t need to manage their own servers, and all enterprise data remains within Microsoft Cloud boundaries, ensuring privacy and security.
