Microsoft Edge is quietly turning the browser into something closer to a working space than a passive window. With its latest Copilot update, Microsoft is giving the AI chatbot a new way to understand what you’re doing: instead of relying only on what you type or what’s currently visible, Copilot can now pull context from across all of your open tabs. That means the assistant isn’t just answering in the abstract—it can reference the material you’ve already gathered, compare what you’re looking at, and stitch together summaries from multiple pages you’ve left open for later.
This is a meaningful shift in how “browser AI” is supposed to behave. For years, the promise has been that AI could help you search faster, summarize pages, or rewrite text. Those are useful features, but they still treat the browser as a collection of separate documents. Microsoft’s new approach treats your open tabs as a temporary workspace—an evolving set of sources that reflect your current intent. If you’ve got product pages open, research articles open, and a couple of comparison reviews open, Copilot can use that entire set to respond more like a knowledgeable collaborator who can see your desk rather than a tool that only reads one page at a time.
What makes the update stand out is not simply that Copilot can summarize content. It’s that it can do so across multiple tabs, and it can support tasks that naturally require cross-referencing. The Verge reports that when you start a conversation with Copilot, you can ask questions about what’s in your open tabs—whether that’s comparing products, summarizing open articles, or answering questions that depend on information scattered across different pages. In other words, Copilot becomes better at the kind of work people actually do in browsers: collecting, comparing, and synthesizing.
To understand why this matters, consider how most browsing sessions unfold. You rarely open a single page and stop. You open several tabs because you’re trying to answer a question that doesn’t live in one place. Maybe you’re deciding between two laptops and you’ve got spec sheets open in one tab, a review in another, and a pricing page in a third. Or you’re researching a topic and you’ve got background explainers open alongside more detailed reporting. Even if you don’t actively switch between tabs constantly, the information is there—waiting for you to connect the dots.
Until now, many AI features have been strongest when you feed them a specific page or a specific snippet. But real decision-making is messy. It involves trade-offs, conflicting claims, and details that only show up after you’ve opened multiple sources. A tab-aware assistant can reduce the friction of that process by letting you ask for synthesis directly, without manually copying text from each tab or repeatedly switching contexts.
Microsoft also emphasizes control. The company says you can “select which experiences you want or leave off the ones you don’t.” That phrasing is important because it signals that Microsoft understands the sensitivity of letting an AI model access your browsing context. People don’t just keep random tabs open; they often include personal information, accounts, work documents, and pages they’d rather not have included in an AI workflow. By allowing users to choose which “tab experiences” Copilot should use, Microsoft is trying to balance usefulness with user agency.
This is where the update becomes more than a convenience feature. It’s a step toward a browser-native assistant that can adapt to different kinds of sessions. For example, you might want Copilot to summarize open articles while you research something, but you might not want it to incorporate certain tabs that contain sensitive information. Or you might want it to compare products you’re viewing, but not to scan everything else you happened to leave open. The ability to opt in—or opt out—helps make the feature feel less like a black box and more like a configurable tool.
There’s also a broader strategic story here. Microsoft has been iterating on Copilot experiences across Windows and the web, and it has been moving toward a model where Copilot is not just a chat window but a layer that can interact with what you’re doing. The browser is a natural place for that because it’s where information is gathered and decisions are made. If Copilot can reliably interpret your open tabs, it can become a kind of “context engine” for everyday tasks: planning, comparing, writing, and learning.
The update arrives as Microsoft also retires Copilot Mode, a feature that previously offered similar tab-aware behavior but with more agentic capabilities. According to the report, Copilot Mode could draw information from your tabs and offered actions like booking a reservation on your behalf. Microsoft has since folded these capabilities into other parts of the product. That retirement matters because it clarifies what Microsoft is prioritizing: rather than focusing on autonomous actions, it appears to be emphasizing a more grounded, conversation-driven experience that helps you reason with your existing sources.
That distinction is subtle but significant. Agentic features can be powerful, but they also raise concerns about reliability and user trust. When an AI takes actions—especially ones that affect money, bookings, or accounts—users need strong guarantees about what will happen and why. A tab-aware chat that focuses on summarizing, comparing, and answering questions can deliver immediate value with fewer risks. It’s easier to validate: you can see the sources, you can ask follow-ups, and you can steer the conversation.
In practice, the new capability could change how people use Copilot during research and shopping. Instead of treating Copilot like a separate destination—something you open to get answers—you can treat it like a companion that understands your current browsing context. That reduces the “tool switching tax,” where you bounce between tabs and copy/paste content into prompts. It also reduces the cognitive load of remembering what you opened. If you’ve already done the legwork of finding sources, Copilot can do the synthesis.
Imagine a scenario: you’re comparing two services and you’ve opened their pricing pages, a policy explanation, and a third-party review. You ask Copilot, “Which one is better for my use case?” A traditional assistant might ask you to clarify your needs or might provide generic advice. A tab-aware assistant can instead ground its response in the specific pages you already opened—highlighting differences in pricing structure, limitations, and real-world commentary from the review. The result is more tailored and less likely to miss details you already found.
Or consider writing. Many people use the browser to collect sources before drafting. If Copilot can summarize open articles across tabs, it can help you build an outline quickly: key points, supporting evidence, and themes. It can also help you spot contradictions between sources. When you’re reading multiple articles, it’s easy to lose track of what each one claims. A synthesis-focused assistant can keep those threads organized.
There’s also a learning angle. Browsing is often how people learn: they open a few explanations, skim a couple of deeper dives, and then try to form a coherent understanding. A tab-aware Copilot can act like a tutor that references the exact materials you’ve chosen. Instead of asking you to re-explain everything, it can ask targeted questions based on what it sees in your open tabs. That can make the learning process feel more interactive and less like passive consumption.
Of course, any feature that pulls information from open tabs raises questions about privacy and boundaries. Microsoft’s mention of selecting which experiences to include suggests it’s thinking about these concerns. Still, users will likely want clarity on what exactly counts as “open tabs,” how long the context is retained, and whether Copilot can access content behind logins or within restricted pages. Even if the feature is designed to be helpful, trust depends on transparency. The best version of this feature is one where users can easily understand what’s being used and can turn it off when needed.
There’s also the question of accuracy. Summarization and comparison across multiple tabs can be extremely useful, but it also increases the risk of mixing up details if the assistant doesn’t handle sources carefully. For example, two tabs might contain similar claims, or one might be outdated. A robust implementation should ideally cite or at least clearly reference which tab contributed which part of the answer. Even if the user interface doesn’t show full citations, the assistant should behave in a way that makes it obvious when it’s drawing from specific pages. Otherwise, users may struggle to verify the output.
Microsoft’s announcement, as described in the report, frames the feature as a way to ask questions about what’s in your tabs and to summarize open articles. That implies the assistant is not just scanning for keywords—it’s interpreting content. The challenge is doing that reliably across different page types: product listings, blog posts, PDFs, dynamic pages, and content that changes depending on location or login state. The more varied the tabs, the harder it is to produce consistent results. If Microsoft gets this right, it will feel like a major leap. If it doesn’t, users will notice quickly.
Still, the direction is clear: Microsoft wants Copilot to be more useful in the moment you’re already working. This is part of a larger trend in consumer AI tools—moving from “ask a question” to “help me complete a task using what I already have.” The browser is one of the richest environments for that because it contains both the user’s intent (what they opened) and the raw material (the content itself). A tab-aware assistant can connect those two.
There’s also a competitive implication. Other browser-integrated AI features exist, but the differentiator here is the breadth of context: across all open tabs, not just the current page. That can make Copilot feel more like it’s participating in your workflow rather than responding to isolated prompts. If users adopt it for shopping comparisons, research summaries, and multi-source Q&A, it could become a default habit—something people rely on every time they open a browser session with a goal.
And that’s where the unique take comes in: this update isn’t just about AI being smarter. It’s about AI being more situational. The assistant is learning to