Google debuted a new “Auto Browse” function for Chrome on Wednesday. The instrument, powered by Google’s present Gemini 3 generative AI mannequin, is an AI agent designed to take over your Chrome browser to assist full on-line duties like reserving flights, discovering residences, and submitting bills.
The discharge of Auto Browse is a part of Google’s continued integration of AI options into Chrome. Final yr, Google dropped the “Gemini in Chrome” mode to reply questions on what’s on internet pages and synthesize particulars from a number of open tabs.
Auto Browse, which customers can entry by launching the Gemini sidebar in Chrome, will solely be accessible at the moment within the US to subscribers of Google’s month-to-month AI Professional and AI Extremely plans. It’s unclear when Auto Browse will turn into accessible to nonpaying customers and extra international locations.
Google’s rollout squares with Silicon Valley’s imaginative and prescient for the way forward for internet searching, which features a complete lot extra AI and a complete lot much less of you. Whether or not it is a browser designed from inception round generative AI, like OpenAI’s Atlas, or one which’s been retrofitted with new AI-based instruments, like Google’s Chrome, nearly each possibility accessible to shoppers now has some stage of baked-in AI. (The Vivaldi browser is a notable exception for customers who need to keep away from AI-powered internet searching.)
In a prelaunch demo, Charmaine D’Silva, a director of product administration for Chrome, confirmed me an instance of Auto Browse serving to her store on-line. “As a substitute of getting to recollect the place I purchased one thing and attempt to reorder one thing,” she mentioned, “I can now delegate to Auto Browse inside Gemini to have the ability to go forward and purchase jackets for me.” By typing a message to the Gemini sidebar in Chrome, D’Silva requested the bot reorder a jacket she purchased final yr, and to discover a low cost coupon code earlier than making the acquisition.
When initiated, Auto Browse takes over Chrome and makes ghostly clicks in its personal tab whereas it makes an attempt to finish the given job. “Use Gemini rigorously and take management if wanted,” reads a disclaimer on the demo model. “You’re answerable for Gemini’s actions throughout duties.” Although you’re sending it off into the digital wilderness, Google nonetheless sees you as answerable for what its bot does on-line whereas following your requests.
For now, the automation solely goes to date. Duties which might be deemed by Google to be extra delicate, like posting on social media and swiping your bank card, nonetheless require a little bit of person oversight. In these conditions, the Chrome bot will lay out the steps it took to get that far in and ask the person in the event that they want to proceed.
Anybody concerned with experimenting with Auto Browse ought to strongly take into account the safety implications of this type of automation. Regardless of Google’s efforts to make it safer to make use of, Auto Browse and comparable AI-based instruments are nonetheless vulnerable to being deceived by prompt injection attacks when visiting malicious web sites, which trick the bot into performing in ways in which the preliminary person didn’t intend.
I’ll be testing Auto Browse this week to get a way of its preliminary strengths, weaknesses, and what the instrument really means for common Chrome customers. Usually, I’m skeptical about agentic AI instruments designed to make your life extra environment friendly and sweep up all of your digital chores. The bots are nearly at all times overhyped, and I’ve discovered them to be persistently unreliable. Nonetheless, Google is insistent on realigning the net searching expertise round AI. Given Google’s observe file for pushing out new options step by step, you must count on Auto Browse to roll out extra extensively within the close to future.
