Although Microsoft has tended to position the Copilot features in Windows more in the background, it continues to develop the standalone Copilot application. Following the company’s more limited refresh earlier this year, the app is now preparing for a more comprehensive interface change. The new version will use a tabbed structure and Microsoft Office applications will integrate more directly with Copilot. In addition, it will be possible to create small applications and workflows with natural language commands. With these changes, Microsoft aims to transform Copilot from just a chat-based artificial intelligence tool into an application that can be used more actively in daily productivity and corporate business processes.
One of the main sections in Copilot’s new interface will be the “Home” tab. Classic artificial intelligence chats of users and general Copilot operations will be carried out in this section. Microsoft states that the Home tab will be able to automatically route user requests between Copilot’s different operating modes in the future. For example, while a short and simple request for information can be answered directly in the chat mode, Copilot’s advanced mode suitable for task delegation can be activated for more complex tasks consisting of more than one step. Thus, the user may not need to choose which Copilot mode is more suitable before each operation. This approach can simplify the user experience, especially by bringing together artificial intelligence tools with different capabilities under a single application.
Copilot will be able to create applications and workflows with natural language
The “Code” tab in the new Copilot application also targets users who do not have programming knowledge. According to the information shared by Microsoft, users will be able to describe the application, tracking tool, dashboard, automation or workflow they want to create in natural language. Copilot will then evaluate the request, select the appropriate method and try to create the relevant solution. Microsoft AI at Work Marketing Director Jared Spataro says that the system will determine the approach itself based on the natural language explanation from the user. As a result, simple desktop tools, small applications for data tracking or dashboards for displaying corporate data can be prepared.
This feature is not intended to replace a classical software development environment. Despite this, it offers a way of using it that can enable employees with limited technical knowledge to be less dependent on developer teams for minor needs. The Code tab may be more practical, especially for users who want to visualize tabular data, regularly track certain information, or support repetitive office processes with small tools. On the other hand, in applications created with natural language, issues such as reliability, data access permissions and how maintainable the resulting code will still require attention. In corporate use, what data these tools can access and how the processes they create are controlled will be as decisive as the convenience provided by the application.
Another section to be added to Copilot, the “Autopilot” tab, focuses on artificial intelligence agents that can work more independently. Thanks to this section, Copilot will be able to continue repetitive tasks without the need for users to give new commands at each stage. The system will, for example, be able to automatically respond to certain correspondence threads or undertake the follow-up of periodic work. In Spataro’s example, Copilot can manage a comprehensive supplier evaluation process from start to finish. Different steps such as creating a calendar, preparing a work plan, preparing for meetings and performing the necessary follow-up operations can be carried out by the system.
In addition, Autopilot will also be able to undertake more interactive tasks, such as requesting updates from people involved in the process. In this respect, the feature offers a more comprehensive structure than classical automation tools that only carry out a certain command. Despite this, the use of such autonomous systems in corporate environments makes topics such as access rights, error control and human approval more visible. The ability of an AI tool to act directly in areas such as meetings, correspondence and business plans can provide efficiency, but the impact of an incorrect or incomplete action can also be broader. Considering that Microsoft wants to make Copilot more useful, especially for businesses, the success of the Autopilot division will depend not only on how many tasks it automates, but also on how predictably and controllably it performs these tasks.
The new structure that Microsoft has prepared for Copilot shows that the application is moving from a chat screen-centered artificial intelligence approach to a broader productivity tool. While the Home, Code and Autopilot tabs focus on different needs, Office integration will enable Copilot to work closer to in-house workflows. Why Copilot will continue to be used as a separate application on the consumer side still remains one of the topics that Microsoft needs to answer. On the corporate side, usage areas such as data processing, report preparation, presentation creation and automating repetitive processes seem more concrete. The practical value of the new Copilot application will largely be revealed by how reliable, easily manageable and consistent these features work in daily work.