Microsoft is testing Copilot Vision, its voice-enabled AI that interacts with your browser

Microsoft has started testing a new feature called Copilot Vision for select Microsoft Copilot Pro subscribers in the US. Introduced in October, this tool allows Microsoft's AI assistant to analyze and provide insights on the content users are browsing in the Microsoft Edge browser, including both text and images. Users can interact with Copilot Vision using their voice to ask questions or request assistance with the content on the webpage, useful for tasks like analyzing Microsoft OneDrive images, interpreting handwritten recipes, or simply answering general questions about the on-screen content.

Copilot Vision will initially work with a limited number of websites, gradually expanding based on user feedback, as Microsoft carefully addresses privacy and security concerns similar to those with the Recall feature tested last month, which officially began testing last month.

The feature currently works only in Edge and is available to a small group of US-based Copilot Pro subscribers in the Copilot Labs program, with plans for broader access in the future.