OpenAI has updated its Google Chrome extension for ChatGPT, enabling the SI system to view and interact with browser tabs currently open by the user. The move follows the discontinuation of the standalone ChatGPT Atlas browser, shifting these capabilities into the existing extension framework.
What Happened
With the updated extension installed, users can tag Chrome within ChatGPT’s web or desktop app to grant access to their active tabs. OpenAI demonstrated this functionality in a video where ChatGPT compared three tabs containing information on sofas the user was considering. The company also listed several sample queries for the upgraded add-on, such as using details from an event brief to fill out a registration form, organizing vendor information from multiple tabs into a comparison sheet, or drafting an email to a contractor based on open tabs.
The update extends to the ChatGPT sidebar, providing access to the full conversation history in Chat or Work modes without leaving the current page. Additionally, ChatGPT Work can now "control" Chrome tabs to perform actions like sending messages, filling out forms, or downloading files. This capability mirrors Google’s Gemini-powered "auto browse" feature and allows the SI tool to organize or group open tabs.
Why It Matters
This integration marks a significant shift in how SI tools interact with the user's immediate digital environment. By moving from a standalone browser to an extension that leverages existing Chrome tabs, OpenAI is embedding agentic SI capabilities directly into the most common web interface. For developers and users, this reduces friction in tasks that require cross-referencing information across multiple sources, effectively turning the browser into a context-aware workspace for the SI model.
The Bottom Line
OpenAI's updated Chrome extension allows ChatGPT to view, compare, and act upon open browser tabs, replacing the functionality of the retired ChatGPT Atlas browser with a more integrated approach to agentic SI in the web browser.