Google officially launched Auto Browse, an “agentic” AI feature for the Chrome browser that allows Gemini 3 to perform multi-step web tasks on your behalf.
The feature transforms Chrome from a passive tool into an active assistant that can navigate websites, click buttons, and fill out forms to handle digital chores like shopping, travel planning, and document organization.
1. Key Capabilities: What “Auto Browse” Can Do
Unlike standard AI that just summarizes text, Auto Browse takes direct action within the browser window.
- Autonomous Shopping: Gemini can identify products from an image (e.g., a photo in your Gmail), search retailers like Etsy or Target, add matching items to your cart, and even find and apply discount codes.
- Form Filling & Administrative Tasks: It can pull data from one tab (like a PDF or a spreadsheet) and use it to populate complex online forms or expense reports on another.
- Travel & Research: You can ask Gemini to “find the cheapest flight to Tokyo in May that fits my calendar,” and it will navigate through Google Flights and airline sites to present you with a final checkout-ready page.
- Subscription Management: Users can instruct the agent to “go through my billing portals and check if all my monthly utilities are paid” or “cancel my trial for this service.”
2. Safety & User Control
Google has positioned Auto Browse as a “monitored automation” tool to address security concerns:
- “Human-in-the-Loop”: The agent is strictly prohibited from making final financial transactions or posting to social media without explicit user approval. It will stop at the “Place Order” button and wait for you to click it.
- Take Over Task: A persistent “Pause” or “Take Over” button sits at the top of the browser, allowing you to intervene at any second if the AI makes a mistake.
- Credentials: With permission, it can use Google Password Manager to log into sites on your behalf, but it proactively asks for consent before accessing sensitive account data.
3. Availability and Limits
As of February 2026, the feature is in a “Preview” phase with specific access requirements:
- Eligible Users: Exclusively available to Google AI Pro ($20/month) and Google AI Ultra ($250/month) subscribers in the United States.
- Daily Task Caps:
- AI Pro: 20 auto-browse tasks per day.
- AI Ultra: 200 auto-browse tasks per day.
- Platforms: Available on Chrome for Desktop (Windows, macOS, and Chromebook Plus).
4. Integrated Experience: The New Side Panel
To support Auto Browse, Google redesigned the Chrome interface:
- Persistent Side Panel: Gemini now lives in a fixed panel on the right side of the screen, allowing you to monitor its progress step-by-step as it navigates your tabs.
- Universal Commerce Protocol (UCP): Google co-developed this new open standard with retailers like Shopify and Wayfair to ensure AI agents can understand and interact with different e-commerce sites seamlessly.
- Nano Banana Integration: Users can also use the Nano Banana image model within the side panel to edit or generate images on any webpage without needing to download them.
