Microsoft has begun testing new features within the Copilot application, in a step aimed at enhancing the integration of artificial intelligence within the Windows environment.
Among the most notable of these features is support for image generation directly within the app, utilizing the GPT-4o model instead of DALL·E 3.
OpenAI had announced last March that GPT-4o now supports image creation and editing capabilities within the ChatGPT platform.
In a related context, a new feature named "Action" appeared within the Labs tab (copilot.microsoft.com/labs), although it has not been fully launched yet.
This specific property allows users to execute short daily tasks on their behalf, and it operates within sessions ranging from five to ten minutes.
The focus appears clearly on the Windows system, and it might initially be made available only to Copilot Pro subscribers.
Correspondingly, the appearance of the virtual characters within the app is changing, especially in voice mode, where they now appear fullscreen.
One of the new characters—which has not yet been named—now possesses clearer features resembling a cloud or a bubble, similar to the "Erin" character whose appearance has evolved gradually.
These updates reflect Microsoft's efforts to integrate AI into productivity, while simultaneously presenting a more interactive and human-like interface.
