ChatGPT’s Chrome extension is becoming more hands-on
OpenAI has added tab-aware capabilities to its ChatGPT extension for Google Chrome, letting the assistant use information from browser tabs a person already has open. The update also expands ChatGPT’s Chrome sidebar and gives ChatGPT Work controls intended for browser-based tasks such as filling forms, sending messages and downloading files.
The change shifts the extension beyond a page-by-page chat tool. Rather than requiring someone to copy material into a prompt, they can tag Chrome in ChatGPT’s web or desktop app and ask it to consider the tabs that are currently open. OpenAI demonstrated the feature with a shopping comparison: a user asks ChatGPT to compare sofas shown across three tabs.
The feature follows the end of OpenAI’s ChatGPT Atlas browser, according to the company’s announcement as reported by The Verge. Instead of maintaining a separate browser product, OpenAI is placing more of its browser-connected functionality inside Chrome through the extension.
What the new tab access is designed to do
OpenAI’s examples focus on tasks in which information is split between websites. A user could ask ChatGPT to extract vendors, prices, key features and source links from several open tabs and place the results into a comparison sheet. Another suggested use is comparing contractor services across tabs and drafting an email to the provider that appears to fit a home-renovation project.
The examples also extend to forms. One proposed request asks ChatGPT to use details from an event brief to fill out a registration form, while stopping before submission. Another asks it to fill out a dentist form and add an appointment to both work and personal calendars. The stated instruction to stop before submission is notable: form completion can be useful, but submitting information or creating calendar events changes a user’s external accounts and commitments.
OpenAI says the updated sidebar provides the full ChatGPT experience without requiring a person to leave the page they are viewing. It can use the user’s full conversation history in ChatGPT or ChatGPT Work. In practice, that means the browser context and an existing conversation can be used together, rather than treated as separate workspaces.
ChatGPT Work adds browser controls
For ChatGPT Work users, the update goes further than reading tabs. OpenAI says the product can control Chrome tabs for tasks including sending messages, filling out forms and downloading files. It can also organize or group tabs.
Those functions place the extension in the growing category of browser assistants that combine information retrieval with actions inside websites. The Verge compared the direction to Google’s Gemini-powered “auto browse” feature. The appeal is straightforward: routine work often involves switching among tabs, copying details into forms or documents, and keeping a set of research pages organized.
But the same capabilities require careful user attention. Open tabs can contain account pages, private correspondence, medical or financial information, internal company documents, or forms ready to submit. A useful workflow is to decide which tabs are relevant before asking the assistant to compare or summarize them, and to review any drafted message, populated form or downloaded file before treating the task as complete.
A browser assistant built around context
The update reflects a broader change in how AI assistants are being positioned. Earlier browser integrations often centered on answering questions about the page in front of a user. Tab awareness allows an assistant to work across a set of sources at once, which is closer to how people perform research, comparison shopping and administrative work in a browser.
That does not make the assistant an independent decision-maker. OpenAI’s own examples are framed as requests to gather information, compare options, prepare a draft or fill a form. The user still needs to determine whether the tabs represent complete and trustworthy information, whether a suggested vendor or contractor is actually appropriate, and whether a form contains the correct details.
For Chrome users who already rely on ChatGPT, the update reduces the friction between browsing and prompting. It also makes the extension a more consequential part of the browser: it can see selected working context and, for ChatGPT Work, take actions in tabs. The practical value will depend on whether users can keep that access deliberate, scoped and easy to supervise.
This article is based on reporting by The Verge. Read the original article.
Originally published on theverge.com
