Google Chrome will get full Gemini integration on Android! It will browse pages and fill out forms for you

Google is taking Gemini in Chrome for Android up a notch. The previous superficial integration, which essentially only managed page summarization via a system overlay, is giving way to full-fledged Gemini directly in the browser. And on top of that, the auto browse feature, which is no longer about answering questions – Gemini will be able to click, scroll, and fill out forms itself to complete a given task for you. The catch is that the sharp launch has a number of conditions.

Gemini moves next to the address bar

Following the example of desktop and iOS, mobile Chrome will get a dedicated Gemini button next to the address bar. Tapping it will slide up a panel from the bottom, where you can ask questions about the current page, have it summarized, or request an explanation of a concept. So, nothing you wouldn’t read in other browsers’ marketing materials, but this time it’s meant to truly operate within Chrome, not as an external overlay.

More interesting is the connection with other Google services. According to Google, from an open page, events will now go to Calendar, ingredients to Keep, and information to Gmail with a single command. In addition, the integration of the Nano Banana model is added, which is designed to generate infographics and other visualizations from text information. How well this will work with Czech content is, of course, a question – Google demonstrates English and refined American services in its examples; local e-shops and Czech websites will be a different discipline.

Auto browse: when Chrome starts clicking by itself

The real highlight of the announcement is auto browse. The feature takes control of the browser interface and attempts to complete a given task itself – scrolling, tapping buttons, filling in fields. Google demonstrates the feature by reserving parking via the SpotHero service based on data from a ticket confirmation email, or by modifying a recurring order on the Chewy e-shop.

To let the user know that Gemini is working, a circular indicator will appear around the icon and the edges of the webpage will glow blue. The status can be monitored directly in the panel, or minimized into a docked progress bar to continue with other activities. For sensitive actions – purchases, social media posts – Chrome will require manual confirmation, which, given the nature of agent behavior, is more a necessity than a friendly gesture.

Who gets it, who waits

Users in the USA will see the rollout from late June. Full Gemini integration will require Android 12 or newer and at least 4 GB of RAM, so the vast majority of active devices should qualify. However, the auto browse feature is reserved for Google AI Pro and AI Ultra subscribers and will only run on unspecified “selected devices.”

Google remains silent on availability in Europe. Given the rules of the AI Act and the generally slower pace of AI feature rollouts in the EU, we are realistically talking about autumn, perhaps the end of the year. And that applies to full Gemini integration – auto browse might wait even longer for Czech users, if it ever arrives in the form Google demonstrated to the American audience.

Would you let Gemini browse the web and handle orders for you?

Source: 9to5Google