Chrome Integrates Gemini AI for Autonomous Task Features
▼ Summary
– Google is adding an AI-powered sidebar to Chrome, integrating its Gemini assistant directly into the browser to compete with new AI browsers.
– The Gemini sidebar can understand context across multiple open tabs, which is useful for comparing products or prices from a single website.
– A new personal intelligence feature will connect Gemini to user data from services like Gmail and Photos, allowing queries about personal schedules or automated email drafting.
– The most ambitious feature is “auto-browse,” an agent that can perform tasks like shopping using personal data, but it will request user approval for sensitive actions like purchases.
– These features, including the sidebar and image modification tools, are rolling out now or soon, though the agent technology faces real-world reliability challenges.
Google is bringing its Gemini AI directly into the Chrome browser with a new persistent sidebar, marking a significant shift in how users interact with the web. This move comes after a wave of new AI-focused browsers entered the market, aiming to challenge Chrome’s dominance with automated assistance. Google’s integration transforms Gemini from a floating window into a constant companion that can understand your open tabs and assist with complex tasks.
The updated Gemini sidebar can now recognize when you have multiple tabs open from a single website, treating them as a related group. This proves incredibly useful for activities like comparing prices or evaluating different products during an online shopping session. Previously limited to Windows and macOS, this sidebar feature is now also rolling out to users of Chromebook Plus devices.
A major advancement is the upcoming integration of Google’s personal intelligence capability. This will connect Gemini to your data from services like Gmail, Search, YouTube, and Google Photos. In the coming months, this means you could ask the sidebar about your family’s schedule or instruct it to draft and send an email without ever opening Gmail. Another new tool, Nano Banana integration, allows for on-the-fly image editing, letting you modify a picture you have with another image or product you find while browsing.
Perhaps the most forward-looking feature is called auto-browse. This agentic function is designed to handle multi-step tasks autonomously by navigating websites on your behalf. You could, for example, ask it to visit a specific online store, purchase an item, and even hunt for a discount coupon. For security, the system is programmed to request user intervention for sensitive actions like logging into a site or finalizing a purchase. Google has stated that these features will utilize Chrome’s existing password manager or saved payment details, but its AI models will not have direct access to this sensitive information. Initially, the auto-browse capability is being released to AI Pro and Ultra subscribers in the United States.
While the potential is vast, browser-based AI agents face real-world hurdles. They can be finicky, sometimes misunderstanding user intent or failing as they move between different websites. Google’s demos, which often focus on shopping and travel planning, highlight the promise but also the challenge of achieving reliable performance for wider adoption. The company reports that in early testing, users have employed the feature for scheduling appointments, filling out lengthy forms, collecting tax documents, obtaining service quotes, and filing expense reports.
The Gemini sidebar support and Nano Banana image tool begin their rollout today. The deeply integrated personal intelligence feature that taps into your Google account data is scheduled to arrive in the coming months.
(Source: TechCrunch)
