ChatGPT’s New Chrome Extension Can Peek at Open Browser Tabs and Control Tasks

OpenAI is bringing computer-use capabilities directly to Google Chrome users with a major update to the ChatGPT browser extension. While the standalone ChatGPT Atlas browser has been discontinued, the newly upgraded Chrome add-on bridges the gap by letting the AI interact directly with a user's web browsing environment.

With the extension installed, users can tag Chrome within ChatGPT's web or desktop application to grant the bot visibility over their open tabs. OpenAI released a demo showcasing the feature in action, where ChatGPT compared multiple open tabs containing different sofa options to help a user make a purchasing decision.

Expanded Sidebar and Task Automation

In addition to tab visibility, the update overhauls the ChatGPT sidebar experience. Users can now access the full power of ChatGPT—including their entire conversation history—directly within a side-by-side interface while viewing any webpage, separating interactions into dedicated "Chat" or "Work" modes.

Through ChatGPT Work, the extension can also actively "control" Chrome tabs to execute multi-step workflows. Similar to Google’s Gemini-powered "auto browse" feature, the updated add-on can perform actions such as:

• Pulling data from event briefs to fill out registration forms (stopping short of final submission).
• Compiling vendor details, pricing, key features, and source links from multiple tabs into a comparison sheet.
• Analyzing open contractor service tabs to draft a targeted email for a home renovation project.
• Filling out online forms ahead of appointments and automatically adding event dates to both work and personal calendars.

The upgraded extension aims to streamline routine web-based productivity tasks by letting users leverage AI contextual awareness across multiple active web pages simultaneously.