Google turns Chrome on Android into an errand-running agent

Google just brought its Gemini browsing assistant to all Android users in the U.S. inside Chrome, and for power users, it is now ready to actually browse for you.

Accessible via the top bar icon or the three-dot menu, Gemini in Chrome summarizes long articles, answers questions about web pages, and connects directly with Google Calendar and Keep without switching tabs. It also lets users generate and customize images on the go using Nano Banana.

Letting AI take the wheel

The real headline is "auto browse," a new agentic feature rolled out to U.S. AI Pro and AI Ultra subscribers.

Instead of just digesting content, auto browse handles tedious everyday chores like booking event parking, updating recurring online orders, or organizing vacation itineraries. To keep things safe, Google says the underlying models are trained to detect prompt injection threats, and the tool asks for confirmation before completing sensitive tasks.

Why it matters

We think this marks a critical shift for mobile browsing from passive consumption to autonomous action.

Basic web summarization and calendar hooks are convenient quality-of-life upgrades, but letting an agent navigate the web and execute tasks on your behalf is where the real value lies. If Google can prove that auto browse reliably deflects prompt injections and avoids booking mistakes, it could turn Chrome from a passive window into an active assistant that genuinely buys back your time.