Google brings agentic AI features to Chrome browser

Google is launching new AI features in Chrome, introducing a persistent Gemini sidebar and agentic capabilities to compete with a growing market of AI-focused browsers. Last year, a wave of AI-centric browsers from companies including OpenAI, Perplexity, Opera, and The Browser Company launched with sidebar assistants and automated features designed to challenge Chrome. While Google introduced its Gemini assistant to Chrome last September, the tool previously operated only within a floating window. The new persistent sidebar allows users to ask questions about active websites or open tabs, and it can group multiple tabs opened from a single page into a context group to help compare products or prices.

Starting today, Google is expanding access to the Gemini sidebar to Chromebook Plus users, whereas the feature was previously limited to Windows and macOS. The update also introduces “Nano Banana,” a new integration that allows users to modify an existing image with another image or product found while browsing. Most notably, Google is rolling out “auto-browse”—an agentic feature designed to traverse websites and perform tasks autonomously—initially to AI Pro and Ultra subscribers in the U.S. The tool can navigate websites to find discount coupons or purchase items, using Chrome’s password manager and saved card details without exposing those details to the underlying AI models. The agent still requires manual user intervention for sensitive actions like logging in or making final purchases.

Google also plans to roll out a “personal intelligence” feature in Chrome in the coming months. This feature connects the AI assistant to personal user data across Gmail, Google Search, YouTube, and Google Photos, enabling users to ask questions about their schedules or draft and send emails directly from the sidebar. However, the wider adoption of these tools faces significant technical hurdles. Browser-based agents are often finicky in real-world use, frequently failing to understand user intent or breaking when navigating across different websites. While Google notes that early testers used the feature for tasks like scheduling appointments, filling out online forms, and filing expense reports, maintaining reliability across the open web remains a major challenge.

Why it matters

By embedding agentic capabilities directly into Chrome, Google is attempting to neutralize the threat from specialized AI startups. The rollout highlights how incumbent tech giants are moving quickly to turn experimental AI agents into default consumer utilities, raising the stakes for independent browser competitors.