Auto Browse is Google’s new AI-powered browser agent integrated directly into the core of Google Chrome. Rather than being a traditional third-party browser extension, it is built directly into Chrome to allow Google’s Gemini 3 AI to navigate the web, click links, and execute complex multi-step tasks on your behalf. Core Capabilities
Form Filling & On-Screen Actions: It reads the page layout to type text, scroll, and click buttons automatically.
Research & Price Comparisons: It can scan multiple open tabs or different vendor sites to compare pricing or pull data.
Travel & Scheduling: Users can ask the agent to find flight and hotel costs across multiple dates or coordinate appointment bookings.
Professional Workflows: It can grab information from a source document and log it straight into corporate CRM systems. How It Works
The Prompt: You trigger Auto Browse by describing a task directly to Gemini in the Chrome side panel.
The Blueprint: The AI reads your open tabs, plans out the necessary steps, and presents the plan for your review.
Execution: Once approved, Chrome opens a dedicated tab with a visual cursor indicator showing the AI interacting with the web pages in real-time.
Human-in-the-Loop Confirmation: For highly sensitive tasks like final checkouts, social media posts, or subscription changes, Auto Browse stops and prompts you for manual approval before finalizing. Availability & Enterprise Features
Consumers: The feature is currently rolling out in the United States for Gemini Pro and Gemini Ultra subscribers on desktop, with an Android mobile version announced as well. There are daily limits on how many agentic tasks it can perform.
Chrome Enterprise: Google also deployed Auto Browse to Workspace users via Chrome Enterprise. This includes a feature called “Skills” that lets users save and reuse repetitive AI workflows across any webpage. Security & Known Limitations Chrome gets new Gemini 3 features, including auto browse
Leave a Reply