When you open Copilot in the Microsoft Edge sidebar, it can automatically read the content of the active browser tab. This feature is called tab context. It allows Copilot to answer questions about the page you are viewing without you having to copy and paste text. Understanding how tab context works helps you control what data Copilot can access. This article explains the behavior, how to enable or disable it, and what limitations exist.
Key Takeaways: Copilot Tab Context in Edge Sidebar
- Edge sidebar Copilot pane > Context toggle: Controls whether Copilot reads the current tab’s content.
- Copilot + Tab key shortcut: Opens Copilot in the sidebar and activates context for the active tab.
- Microsoft Edge > Settings > Privacy, search, and services > Copilot: Manage global permissions for tab context across all sites.
How Copilot Tab Context Works in the Edge Sidebar
Copilot in the Edge sidebar can access the visible text content of the active browser tab. This includes the main article content, headings, and visible text. It does not read behind logins, payment forms, or password fields. The feature is designed to help you summarize pages, ask questions about specific content, or extract key points without manual copying.
Tab context is not enabled by default for all users. It requires a Microsoft 365 subscription for Copilot Pro or Copilot for Microsoft 365. Free Copilot users in Edge can still use the sidebar but must manually provide context by copying text into the prompt. The toggle for tab context appears only when a compatible subscription is detected.
When tab context is active, Copilot processes the page content on Microsoft servers. The content is not stored permanently. Microsoft states that the data is used only to generate the response and is not used to train the underlying models. This is similar to how Copilot in other Microsoft 365 apps handles grounded data.
What Tab Context Can Read
Copilot reads the rendered text of the webpage. This includes:
- Article body text
- Visible headings and subheadings
- List items and table content
- Visible alt text on images
It cannot read:
- Content inside iframes from different domains
- Password fields or masked input fields
- Dynamic content loaded after the page is rendered, such as infinite scroll sections not yet visible
- PDF files opened in the browser unless embedded as text
Enabling and Disabling Tab Context in the Edge Sidebar
You control tab context through the Copilot pane and Edge settings. The toggle is session-based. It resets each time you close and reopen Edge unless you change the global setting.
- Open the Copilot pane in Edge
Click the Copilot icon in the Edge toolbar or press Ctrl + Shift + period on Windows. The pane opens on the right side of the browser window. - Locate the context toggle
At the top of the Copilot pane, look for the toggle labeled “Allow Copilot to access page content.” The exact label may vary by update. It is usually a switch with the text “Page context” or “Allow access to page content.” - Toggle the switch on or off
Set the toggle to the On position to let Copilot read the current tab. Set it to Off to prevent Copilot from reading the tab. The change applies immediately to the current session. - Set a default behavior in Edge settings
Go to Edge Settings > Privacy, search, and services. Scroll down to the Services section and click Copilot. Under “Page context,” choose “On” or “Off” for the default. This setting persists across browser restarts.
Common Misconceptions and Limitations
Tab context works only on the active tab
Copilot reads only the tab that is currently visible in the browser window. If you switch to another tab, the context changes to that new tab. Copilot does not aggregate content from multiple tabs. To use content from multiple pages, you must copy text manually from each tab into the prompt.
Tab context does not work on all websites
Some websites block automated reading of their content. Copilot’s tab context may not function on sites that use strict Content Security Policy headers or that require authentication. In those cases, the toggle may appear on but Copilot returns a message stating it cannot access the page content.
Tab context does not remember previous pages
Each new conversation in Copilot starts fresh. Tab context applies only to the page you are viewing at the moment you send a prompt. If you navigate away from the page and then ask a question about it, Copilot no longer has access to that content. You must navigate back to the original page for Copilot to read it again.
Copilot Tab Context vs Manual Context Input
| Item | Tab Context Enabled | Manual Context Input |
|---|---|---|
| Description | Copilot reads the active tab content automatically | You copy and paste text into the prompt |
| Ease of use | No extra steps required | Requires manual selection and copying |
| Supported content | Visible rendered text only | Any text you can copy, including parts of PDFs |
| Session persistence | Resets when tab changes or browser closes | Persists in the conversation until you clear it |
| Subscription needed | Copilot Pro or Copilot for Microsoft 365 | Works with free Copilot in Edge |
Choose tab context when you need quick answers about a single page. Use manual input when you need to combine content from multiple sources or when the page content is not accessible to Copilot.
You can now control exactly when Copilot reads your browser tab content. Use the toggle in the Copilot pane for per-session control or the Edge settings for a persistent default. For sensitive pages such as banking or personal email, disable tab context before opening Copilot. This ensures no page content is sent to Microsoft servers during your session.