Google is deepening its push into AI-driven commerce and autonomous browsing with two major updates aimed at reshaping how users discover and purchase products online. The company has introduced a new shoppable discovery feed inside Doppl, its experimental virtual try-on app. The feed uses AI-generated videos of real merchandise to recommend outfits that match a user’s personal style, based on their interactions in the app. Nearly every item includes a direct link to a retailer, supporting a seamless path from discovery to purchase. Google sees this format as a way to compete more effectively with visual commerce leaders like Amazon, TikTok, and Instagram, while offering a customizable virtual try-on experience. The feature is now rolling out to iOS and Android users in the U.S.
At the same time, Google is preparing Chrome to safely support more autonomous, agent-driven actions such as booking, shopping, and form-filling. The browser now uses oversight models, including a Gemini-powered evaluation system, to confirm actions align with user intent and restricts access to sensitive web content through origin-based permissions. User approval remains mandatory for high-risk tasks like payments or navigation to secure websites. Together, these updates reflect Google’s strategy to merge convenience with control as AI becomes more active in everyday decision-making.
