Autonomous Browsing Agents

Google has introduced an Auto Browse feature for its Chrome browser that uses the company’s Gemini 3 generative AI model to carry out web tasks on a user’s behalf. Available initially in the United States for subscribers to Google’s AI Pro and AI Ultra plans, the agent is designed to handle online chores such as booking travel, searching for rentals, or managing expenses. The core differentiator is its ability to actively operate the browser, rather than just summarizing pages or answering questions.

Auto Browse is accessed through the Gemini sidebar in Chrome, where users type instructions and then watch as the agent takes over a tab and performs clicks and navigation steps. In demonstrations, the tool reordered an item from a past purchase and searched for discount codes before checkout. For higher-risk actions like social posts or payments, the system pauses and presents a step-by-step summary, asking the user to confirm before proceeding.

For consumers, Auto Browse reflects a shift from passive AI helpers toward agentic tools that execute multi-step workflows online. This may streamline repetitive digital tasks for busy users already embedded in Google’s ecosystem, particularly around shopping and trip planning. The feature also aligns with a broader trend in browsers, where AI is becoming a built-in layer for task completion rather than an optional add-on, signaling a new phase in how people navigate and interact with the web.

Image Credit: Google