Chrome for Android is becoming an autonomous web agent next month, and Google is betting enterprises will pay for the privilege. At its Android Show: I/O Edition event on Tuesday, Google unveiled Gemini Intelligence, a platform-wide AI layer that turns the browser from a passive content viewer into an active agent that navigates websites, fills forms, and completes transactions on a user's behalf.
The feature, called auto browse, lets Chrome act on instructions across the open web, reserving parking spots, updating recurring orders, or copying grocery lists from notes into shopping carts. The catch: auto browse is locked behind Google's AI Pro ($20/month) and AI Ultra ($250/month) subscriptions and limited to US devices running Android 12 or higher at launch.
Built on Gemini 3.1. The update brings three capabilities to Chrome for Android at the end of June. A persistent AI assistant inside the browser can summarize articles, explain content, and pull data from connected Google services like Gmail, Calendar, and Keep. A tool called Nano Banana generates and edits images directly in the browser. But auto browse is the headline, it's the first time Google has let a browser act as an autonomous agent on mobile, taking multi-step actions across the web without the user driving each click.
Google has built in a confirmation step for sensitive actions like purchases or social media posts. But the agent operates with the same permissions as the user, meaning it can read Gmail inboxes and act on the contents. For IT and security teams managing fleets of Android devices, that changes the risk calculus.
Prompt injection, where malicious content on a webpage hijacks the AI's instructions, remains one of the hardest unsolved problems in agentic AI security. Google says auto browse is protected against it, but organizations browsing supplier portals, financial platforms, or customer-facing tools will want to verify those protections before the feature lands on managed devices. The enterprise implications extend beyond security. Google's Personal Intelligence feature lets Gemini tailor responses based on a user's interests, family details, and browsing habits by pulling context from Gmail, Google Photos, and Calendar to pre-populate forms.
Google says it's opt-in and users can disable it via settings, but privacy teams will need to examine what data is used, where it's processed, and how it interacts with corporate Google Workspace accounts.
Gemini Intelligence also introduces Rambler, a Gboard dictation feature that strips filler words and handles mid-sentence language switching. And Create My Widget lets users build custom home screen widgets from natural language descriptions, similar to Nothing's Essential Apps tool from last year. The rollout begins with Samsung Galaxy S26 and Google Pixel 10 devices this summer, with broader Android distribution following later in 2026. Beyond phones, Google plans to extend Gemini Intelligence to Wear OS, Android Auto, Android XR, and a new Googlebook laptop category from Acer, Asus, Dell, HP, and Lenovo arriving this fall. For IT teams, the immediate action is straightforward: flag this to security and compliance stakeholders before auto browse appears on employee devices. The browser is no longer a neutral piece of infrastructure, it's an agent with the keys to the user's accounts, and the policies that governed it yesterday don't apply today.
