ChatGPT Can Now See Your Chrome Tabs — How OpenAI’s New Extension Works

OpenAI is done asking you to switch browsers. The company has quietly retired its standalone Atlas browser experiment and replaced it with something far more practical: an upgraded ChatGPT extension for Google Chrome that lets the assistant see — and act on — the tabs you already have open. Announced on 15 October 2026, the update brings OpenAI’s computer-use capabilities to the browser most of the world already uses.

The mechanics are straightforward. With the extension installed, you can tag “Chrome” in a ChatGPT prompt from the web or desktop app, and the model will pull content from your open tabs into the conversation. OpenAI’s demo showed a user asking ChatGPT to compare three sofas, each open in a separate tab, on dimensions, materials, delivery times and return policies — the kind of cross-tab comparison that previously required tedious copying and pasting.

Chat mode and Work mode

The feature splits into two modes. In Chat mode, ChatGPT looks across your open tabs and gathers information: comparing products, summarising articles, checking pricing pages against your requirements. In Work mode, it goes further and can actually control your Chrome tabs — sending messages, filling out forms, downloading files — as well as organise or group the pile of tabs you have accumulated.

OpenAI’s sample prompts illustrate the ambition. Users can ask the assistant to put vendors from open tabs into a comparison sheet with pricing, key features and source links; to compare contractor services across tabs and draft an email to the best fit for a home renovation; to extract details from an event brief and fill out a registration form without submitting it; or to prepare a dentist registration form and add the appointment to a calendar. The point is to eliminate the copy-paste shuffle between browser and chatbot.

The sidebar that stays with you

Alongside tab access, OpenAI has refreshed the extension’s sidebar: a persistent ChatGPT panel that sits beside the page you are viewing, with the full power of the assistant and your entire conversation history available without leaving the page. According to the Chrome Web Store listing — version 1.26.1005.11759, updated 15 October, reporting 7 million users — the extension lets ChatGPT pull in context and take action in tabs, with access to installed plugins, full conversation history and the local file system if connected.

Crucially, the listing states that ChatGPT can work inside the browser in the background without taking over the active browsing session, and that it will always ask for permission when referencing browser history and when downloading or uploading files. That permission gate is the key privacy control in the design — the moments of highest risk are the moments the system promises to ask first.

The privacy question

Letting an AI assistant see your open tabs is a significant trust decision. Tabs are a rich source of personal and professional data: banking sessions, medical searches, work documents, private messages. The design, as described, respects user agency — the capability activates only when you explicitly tag Chrome in a prompt, so it is not passive surveillance. But the prompt is the only gate, and users will need to be deliberate about what they expose.

OpenAI’s stated safeguards — permission prompts for browser history, downloads and uploads; background operation that doesn’t hijack the session — address the most obvious risks. What remains less clear is how tab content is stored, how long it is retained, and how it interacts with the company’s data-retention and training policies. Users handling sensitive material should check OpenAI’s current documentation and consider the extension’s data controls before granting access. Enterprise and workspace settings may further restrict what the extension can do, and administrators should review those controls before rolling it out.

The Atlas pivot

The strategic context matters. OpenAI’s Atlas browser asked users to adopt an entirely new browser — a high-friction request in a market where Chrome holds a dominant share. The extension removes that friction completely. Rather than competing with Chrome, OpenAI is now riding on top of it, which is a far more efficient distribution channel for embedding its assistant into daily workflows.

The move also answers Google directly. OpenAI itself compares the feature to Google’s Gemini-powered “auto browse” capabilities for organising and grouping tabs — an acknowledgement that the browser-assistant race is now being fought inside Chrome itself, not in a separate app. For OpenAI, winning means being present where users already are; for Google, it means defending home turf with its own assistant deeply integrated into the browser it owns.

Beyond Chrome

Despite the branding, the extension is not Chrome-only. OpenAI documents tab mentions, side chat and browser control across Chrome, Edge, Brave and Vivaldi, with Opera supporting tab mentions and browser control from the desktop app but not side chat. Availability varies by product rollout and workspace settings, but the strategy is clear: meet users in whichever Chromium browser they already use, and make the assistant feel like a native part of it.

What it means for how we work

For knowledge workers, the practical impact is a quiet but real change in digital multitasking. Research that once meant juggling a dozen tabs and a spreadsheet — comparing vendors, prices, contractors, policies — can now begin with a single prompt. The efficiency gain is genuine, and it raises the bar for what counts as focused work when the mechanical parts of comparison can be delegated to software.

The trade-off is the trust boundary. Every convenience here is purchased with visibility into your browsing. For routine shopping and research, most users will find that an easy trade. For anything sensitive, the deliberate, opt-in design means the safest approach is also the simplest: only tag Chrome when you mean it.

FAQs

How do I let ChatGPT see my Chrome tabs?

Install or update the ChatGPT extension for Chrome (version 1.26.1005.11759 or later), then tag “Chrome” in a prompt from ChatGPT’s web or desktop app. The assistant can then read your open tabs to gather information in Chat mode, or take actions like filling in forms in Work mode. The feature also works in Edge, Brave, Vivaldi and Opera, with some differences between browsers.

Does ChatGPT watch my tabs all the time?

No — access is opt-in per prompt. The capability activates only when you explicitly tag Chrome, so it is not passive surveillance. OpenAI also says ChatGPT always asks for permission when referencing browser history or downloading and uploading files, and that it can work in the background without taking over your active session.

What can it actually do with my tabs?

In Chat mode it compares and summarises across tabs — products, prices, articles and requirements. In Work mode it can control tabs: filling out forms (stopping before submission if you prefer), sending messages, downloading files, building comparison sheets from multiple vendors, and organising or grouping open tabs.

What happened to OpenAI’s Atlas browser?

OpenAI has quietly retired the standalone Atlas browser experiment. The Chrome extension is its replacement strategy: instead of asking users to adopt a new browser, OpenAI is embedding ChatGPT into the browser people already use — a much lower-friction path to the same goal of an AI assistant inside daily browsing.

Is it safe to use with sensitive tabs open?

Exercise caution. Your tabs may contain banking, health or work data, and the prompt is the only gate between that data and the model. OpenAI’s permission prompts cover history access and file transfers, but you should review the extension’s data-retention settings and your workspace controls first — and avoid tagging Chrome when sensitive pages are open.