Back to comparisons
NewsLearn & research

ChatGPT’s Chrome extension gains tab context and browser actions

The Decoder reports that ChatGPT can now use open Chrome tabs as context and perform browser actions from Work mode. OpenAI’s official product materials confirm the broader sidebar and browser-control direction, while some rollout details remain documented only in the report.

Open browser pages feed a side conversation, while a separate gate controls page actions.
CurioFuse · AI-generated illustration
In this article

Open tabs become usable context

The Decoder reports that ChatGPT’s updated Chrome extension can read context from tabs already open in the browser. Users can point a conversation at Chrome through an @chrome mention, while Chat mode can combine information from several tabs. The article says the full ChatGPT conversation history is available in a sidebar beside the current webpage.

The practical examples are comparison and synthesis tasks: matching products across tabs by dimensions, materials, delivery time and return policies; explaining several articles in simpler language; or comparing pricing pages from different sellers. According to The Decoder, the updated extension is available across ChatGPT plans.

Work mode can operate the browser

The report distinguishes reading from action. In Work mode, ChatGPT can use Computer Use to control browser tabs, fill forms, send messages, download files and group tabs. These capabilities can turn a collection of already open pages into a working context rather than requiring the user to copy links and text into a separate conversation.

OpenAI’s browser-extension guide directly documents open-tab mentions, page context and a side chat in Chrome. It also describes starting browser tasks from the desktop app in Work or Codex. An August 31 release-note entry already mentioned tab context and browser control, so the October report should not be read as the first appearance of every capability.

Availability and control still matter

The Decoder’s short report attributes the newest open-tab details to ChatGPT communications on X. At the time of checking, OpenAI’s public release-notes page did not yet contain a separate October 16 entry describing @chrome, multi-tab Chat mode or plan-by-plan availability. Those rollout details should therefore be treated as reported behavior rather than a fully documented support matrix.

Browser assistance also does not mean unrestricted automation. OpenAI’s official materials emphasize that users decide which websites and connected tools ChatGPT may access, and that important actions can require confirmation. Workspace settings, client version, account plan and the availability of the browser extension can also affect what is usable.

The update’s main significance is architectural: ChatGPT is moving from a separate destination toward a companion that can understand the pages already in view and, in an explicitly agentic mode, act on them. For users, the distinction between summarizing tabs and performing consequential actions remains important, especially when forms, messages, downloads or signed-in services are involved.

Sources(1)

Sources & standardsReport a correction