Microsoft is testing its own AI browser via Copilot Mode for Edge

From today, Microsoft is launching Copilot mode for Edge, an experimental version of Edge which places the front co -pilot and in the center of the new tabs and allows him to see the content on all your open tabs.
Tip-toeing in this space, Microsoft describes Copilot mode for Edge as “experimental” and “free for a limited time”, suggesting that it will eventually be part of a paid subscription. The limits of use will apply, although Microsoft does not say what they are still.
Microsoft also says that Copilot mode is optional, opt-in and can be returned or deactivate via the setting panel at any time.
Light it, however, and the “new tab” page-which generally imitates the Windows Widgets panel by new and other content provided by MSN and other publications-will be replaced by a simple co-pilot box like the above, similar to the box on the Copilot application itself.

Microsoft
What Copilot mode seems to offer is Microsoft’s official response to Navigation efforts on Opera, Openai, Perplexityai and others, where an AI chatbot feeds a user’s search request, rather than presenting a traditional links. Microsoft also displays COPILOT mode for Edge as a window of his first foray into agency AI, where “independent” AI agents come out and carry out tasks for you, independently. Finally, a Microsoft demonstration video indicated that it could search and clean the information selected in a video, rather than forcing you to watch all of the content.
All this can be user -friendly, but Microsoft also demonstrates insensitivity to creators that he had not shown before. In a video demonstrating the co-pilot mode for Edge and how he could analyze a web page for a cookie recipe, the invite read: “Can you jump the history of life and arrive at the recipe?”

The prompt has clearly referred to the tendency of recipe creators adding additional details to provide a context and improve their search ranking, but the co -pilot prompt relegated it to the cold efficiency of an order for quick training. This is similar to the way Google’s IA mode now sees the content: something to be exploited, not presented.
Copilot’s ability to “see” all your available content grows further on the way that Windows’s Copilot vision has already taken. Copilot Vision can already “see” the content of two applications that you can choose to share with it, and Microsoft also tests the possibility of “seeing” the entire desk. Copilot mode will extend this to several tab pages – Microsoft does not specify a limit, nor the number of windows to which it will apply. Microsoft has also said nothing about the exclusion of open tabs in private or incognito mode.
We don’t know how fast Microsoft will go to agent navigation. An example of a demonstration asked Edge to compare several tabs, to search for the weather at some point of the day, then to reserve a paddleboard reservation, similar to what was observed in Microsoft’s shopping demonstration during its birthday celebration. As a rule, these transactions force the user to disconnect from the final agreement.

Microsoft
A representative of Microsoft said that the COPILOT mode “will scan” the transcription of a video, so that it can rub against a particular segment that interests you. Microsoft also praises the capacity of Copilot mode to suggest the next steps to perform a particular task, but that already does it.
Finally, Microsoft says that it will want Copilot mode to access your browser history and related data to better respond to your requests and select new directions and tasks. However, it remains to be seen if users will be likely to see.

