Chrome takes on AI browsers with tighter Gemini integration, agentic features for autonomous tasks

Over the past year, a wave of AI-powered browsers from companies such as OpenAI, Perplexity, Opera, and The Browser Company has emerged, all aiming to challenge Chrome with features like built-in assistants, sidebars, and automated task handling. Now Google is responding by expanding its own AI capabilities inside Google Chrome, the world's largest browser by market share.

Google first brought its Gemini assistant to Chrome in September, but the feature was limited to a floating window. With this latest update, Gemini is moving into a persistent sidebar, allowing users to ask questions about the website they are currently viewing or about other open tabs without leaving the page.

One of the features Google previewed ahead of the rollout focuses on how Gemini understands multiple tabs. When users open several tabs originating from the same webpage, the Gemini sidebar treats them as a single context group. This approach is designed to make tasks such as comparing prices or evaluating similar products easier, especially when shopping online.

Image Credits: Google

Previously, Gemini in Chrome was available only to users on Windows and macOS. As part of this rollout, Google is also extending the Gemini sidebar to Chromebook Plus users, broadening access across its ecosystem.

Google is also planning to integrate its recently announced personal intelligence feature into Chrome. This capability connects Gemini to individual services such as Gmail, Search, YouTube, and Google Photos, enabling the assistant to answer questions using a user's own data. Once this feature arrives in Chrome in the coming months, users will be able to ask Gemini about things like family schedules or have it draft and send emails directly from the sidebar without opening Gmail.

Another addition on the way is a new Nano Banana integration, which will allow users to modify existing images by combining them with other images or products they encounter while browsing the web.

The most ambitious update, however, is a new agentic feature called auto-browse. This tool is designed to perform tasks on a user's behalf by navigating websites using the user's personal information. For example, users could ask the agent to visit a specific site, purchase an item, and look for a discount coupon. Google said the system will request user confirmation for sensitive actions, such as logging into accounts or completing purchases.

Last year, Google said these kinds of features would rely on Chrome's built-in password manager and saved payment details, while ensuring that its AI models do not have direct access to that sensitive information.

Auto-browse is initially rolling out to AI Pro and Ultra subscribers in the United States.

Browser-based agents remain challenging because they often struggle with complex real-world tasks. Google's demonstrations, similar to many AI demos, focused on shopping and travel planning. In practice, such agents can misunderstand intent or fail while navigating multiple websites, which could limit broader adoption.

Despite those challenges, Google said early testers have already used auto-browse for tasks including scheduling appointments, completing lengthy online forms, gathering tax documents, requesting quotes from plumbers and electricians, and filing expense reports.

According to the company, Gemini's sidebar support and the Nano Banana integration are rolling out today. At the same time, the personal intelligence features will arrive in Chrome in the coming months.