Chrome becomes an "agent": Google's new AI that independently manages purchases and procedures -

The most ambitious update in browser history turns your browser into a standalone assistant.From managing tax documents to arranging travel and redesigning images, AI is now taking control.

Chrome becomes "active": Google integrates Gemini 3 to automate tasks and browse without tabs

Google today announced a paradigm shift in web browsing.The latest Chrome update isn't just an aesthetic upgrade;it's a deep integration of Gemini 3, the company's most intelligent and powerful AI model to date.

This development turns the browser into an "agent" tool that can perform complex actions on behalf of the user, marking a turning point for MacOS, Windows, and Chromebook Plus.

Complete jump between tabs: a new multitasking experience

The company has introduced a completely redesigned navigation bar that acts as a constant navigation assistant.The thing is simple: eliminate friction when switching between tabs.

This panel not only answers questions, but also contextualizes navigation.The company's testers are already using it to compare options between dozens of open tables, summarize extensive product reviews on different websites, and find free space in extremely busy tables while keeping the main desk intact.

Creativity without downloads: intelligent image insertion

In terms of creativity, this update stands out due to the integration of Intelligent Image Insertion, with the Nano Banana model.This feature allows you to convert and generate visual content in real time right in the browser window.

Unlike traditional tools, users don't need to download, edit and re-upload images. Just enter instructions in the side panel and you can instantly redesign interior spaces, convert sketches or turn research data into high-quality infographics.

Full integration with "Integrated Programs"

Synergy with Google's ecosystem reaches a new level of depth.Through Related Applications, Gemini in Chrome sourced and works with Gmail, Calendar, YouTube, Maps, Google Shopping and Google Flights.

A practical use case is logistics planning: AI can track an old email with conference details, check available flights on Google Flights, check the calendar agenda, and automatically send an email to colleagues informing them of the exact time of arrival.

"Browse automatically 2": jump to task delegate

The most disruptive innovation is undoubtedly Chrome Auto Browse 2.This feature goes beyond traditional autofill to provide an agent-like experience.Initially available to AI Pro and Ultra customers in the United States, it allows you to delegate tasks that would typically require hours of manual browsing.

- Personal logistics: research hotel and flight costs by combining different dates to optimize your budget.

- Professional management and processes: schedule appointments, fill out online forms, collect tax documents, request quotes from service providers and manage driver's license renewals.

- Smart shopping: Thanks to Gemini 3's multimodal capabilities, the system can recognize items from an inspiration image (such as furniture or clothes), search for them in stores, use discount codes and add to the shopping cart.

- Access management: with the user's permission, the tool can use Google Password Manager to navigate and complete tasks even on websites that require login.

New Global Standard: UCP (Universal Commerce Protocol)

In order to ensure that these machines on the Internet are free of any problems, Google has implemented the Universal Commerce Protocol (UCP).It is an open commerce standard developed in partnership with industry leaders such as Shopify, Etsy, Wayfair, Target and Walmart, and is supported by Visa and Mastercard.

The goal is for AI agents to be able to work on behalf of the user in a consistent and error-free manner on any shopping platform.

Security and the "Alignment Critic"

AI independence comes with strong controls.Google has implemented a 'User Alignment Critical Model', a process that evaluates AI services to ensure they actually match the user's intent.

In very sensitive situations, such as making a final payment, managing a purchase or posting to social media, automatic browsing is designed to stop and ask for explicit confirmation.We've built Gemini protections into Chrome to protect against new types of online threats," the company stressed, stressing that the user always retains ultimate control.