Microsoft has launched a new AI-powered feature for its Edge Browser. Copilot Vision is now in preview for a limited number of U.S. Copilot Pro subscribers by opt-in through Copilot Labs. With user permission, Copilot Vision “sees” what is onscreen and can respond to questions about text and images, explains the company. Calling Copilot Vision “the first AI experience of its kind,” Microsoft suggests the experience is “almost like having a second set of eyes as you browse,” adding that when users turn on Copilot Vision it will “instantly scan, analyze, and offer insights based on what it sees.”

Tom’s Guide calls it “a major leap forward in browser-based AI assistance,” making the web experience “more personal and dynamic.”

A Microsoft blog post shows example of users interacting with Copilot Vision as it helps plan a museum trip, help with holiday shopping and assists with learning a new game.

The Copilot Pro plan that includes access to Copilot Labs, thus potentially Copilot Vision, costs $20 per month.

“Beyond answering questions, Copilot Vision can summarize and translate text, and handle tasks like spotlighting discounted products in a store catalog,” writes TechCrunch, adding that “it can also serve as a game assistant, for example offering pointers during matches on Chess.com.”

When Copilot Vision is enabled, “it sees the page you’re on, it reads along with you, and you can talk through the problem you’re facing together,” Microsoft says. “It’s a new way to invite AI along with you as you navigate the web, tucked neatly into the bottom of your Edge browser whenever you want to ask for help.”

Microsoft also emphasizes security. “Once you end a session with Vision, all data about what you say and the context you share with Copilot is deleted,” the company states, vowing “every time you use Vision, your data is handled and protected in accordance with the Microsoft Privacy Statement.”

Copilot’s responses are, however “logged to improve our safety systems,” Microsoft says, stressing that “Vision does not capture, store or use any data from publishers to train our models” and that the company is “prioritizing copyright, creators, and our user’s privacy and safety — and putting them all first.”

For the preview, Vision will only interact with a select set of websites, The Verge points out, adding that Microsoft intends to “cautiously expand this list” over time.