Google is quietly redefining what a web browser does.
With the launch of Auto Browse, a new AI-powered feature in Chrome, Google is moving beyond search and tabs into something closer to a digital assistant that can do things for you. Powered by the Gemini 3 model, Auto Browse can handle multi-step online tasks—researching, comparing, filling forms, logging in, and even checking out—while you watch or step away.
The feature is currently limited to AI Pro and Ultra subscribers in the US, but its implications stretch far beyond a single rollout.
What is Google Chrome’s Auto Browse feature?
Auto Browse is a task automation tool built directly into the Chrome browser and powered by Google’s Gemini 3 AI model.
Instead of helping users find information, Auto Browse is designed to complete entire workflows across websites. That includes actions that normally require constant clicking, copying, pasting, and decision-making.
Think of it less as a chatbot and more as a browser-native assistant that can operate the web on your behalf.
Google positions Auto Browse as especially useful for repetitive, time-consuming tasks—things users know what they want to do but don’t want to spend time doing manually.
What kinds of tasks can Auto Browse handle?
According to Google and early testers, Auto Browse can handle a wide range of everyday online tasks.
Common use cases include:
- Researching travel costs across multiple sites
- Booking appointments or services
- Filling out long online forms
- Managing subscriptions
- Shopping and price comparison
- Collecting documents for taxes or reimbursements
Testers have reportedly used it to:
- Schedule medical or service appointments
- Get quotes from plumbers and electricians
- Track down and download tax-related documents
- Handle repetitive government or utility forms
For vacation planning, for example, Auto Browse can research flights, compare hotels, apply filters, and stay within a preset budget—without the user having to open dozens of tabs.
How does Auto Browse actually work inside Chrome?
Auto Browse operates through Gemini in Chrome, Google’s experimental AI layer embedded in the browser.
Once activated, users can describe a task in plain language, such as “Find me a flight under $800 and book it,” and Gemini attempts to carry out each step across the web.
Importantly, Auto Browse doesn’t just scrape information. It can:
- Navigate websites
- Click buttons and menus
- Enter data into forms
- Move between pages
- Complete checkouts (with user permission)
This places it closer to an AI agent than a typical assistant.
What makes Auto Browse different from earlier AI tools?
The standout feature is multimodal capability.
What multimodal means here
Auto Browse can understand and act on:
- Text
- Images
- Page layouts
- Shopping visuals
For example, it can:
- Identify items from photos
- Search for visually similar products
- Add items to a cart
- Apply discount codes automatically
- Keep purchases within a defined budget
This is a major leap from text-only AI tools, which often struggle with real-world web interfaces.
Can Auto Browse log into websites and make purchases?
Yes—but only with explicit permission.
If users allow it, Auto Browse can:
- Use Google Password Manager to securely sign in
- Complete tasks that require authentication
- Proceed through checkout flows
Google emphasizes that credentials are handled through existing password infrastructure, not stored directly by Gemini.
Still, this is where Auto Browse crosses into sensitive territory, because mistakes now carry real-world consequences.
How to access Auto Browse in Chrome
Auto Browse isn’t available to everyone, at least not yet.
Eligibility requirements:
- You must be 18 or older
- You must be a US resident
- You need the latest version of Chrome
- You must subscribe to Google AI Pro or AI Ultra
- You must opt into Gemini in Chrome
- You must be signed in with a personal Google Account
Important limitations:
- Not available in Incognito mode
- Not supported on work or school accounts
This controlled rollout suggests Google is still testing how users interact with autonomous browser behavior.
What are the risks of using Auto Browse?
Google is unusually direct about this: users are responsible for what Auto Browse does.
Google’s warnings include:
- Gemini in Chrome is still experimental
- Auto Browse may make mistakes
- Purchases, bookings, or errors are the user’s responsibility
- Users should supervise sensitive tasks
To maintain transparency:
- Users can review or delete interactions in the Gemini Apps Activity
- Automated browsing sessions appear in Chrome History, marked with a special icon
This is Google acknowledging a key challenge of AI agents: autonomy without accountability is dangerous.
Why Auto Browse signals a bigger shift in how we use the web
Auto Browse isn’t just a new Chrome feature; it’s a preview of how Google sees the future of browsing.
Instead of humans adapting to websites, websites are increasingly being adapted for AI agents.
This raises bigger questions:
- Will websites start designing for AI visitors, not humans?
- What happens to ads, affiliate links, and comparison shopping?
- How much autonomy is too much when AI handles money and credentials?
Auto Browse suggests Google believes the browser itself—not search, not apps—will become the control center for AI-driven tasks.
TL;DR
- Google has launched Auto Browse in Chrome, powered by Gemini 3
- It automates multi-step tasks like shopping, bookings, and form-filling
- It includes multimodal features for images, carts, and budgeting
- The tool can log into sites and complete purchases with permission
- It’s currently limited to AI Pro and Ultra users in the US
- Google warns users they’re responsible for any mistakes
