Google's Gemini Spark can now do your web chores in Chrome

Looking for flights or trying to schedule an apartment viewing usually means clicking through dozens of repetitive web pages. Google wants its AI to do the boring work for you.

Google just integrated its AI agent, Gemini Spark, directly into Chrome with a new feature called Chrome auto browse. At the same time, the company expanded Spark access to Google AI Pro subscribers in over 160 additional countries.

Running web errands with your saved logins

With your permission, Spark can now tap into your logged-in accounts and saved Chrome passwords. It can handle multistep tasks like researching flights and kicking off the booking process, or locking in a viewing time for an apartment you saved earlier.

To keep things safe, Google built in defenses against prompt injection attacks. Sensitive actions like final payments are automatically handed back to you so the AI never spends money without your sign-off. For now, the auto-browse capability is rolling out exclusively in the U.S. before hitting other regions.

Why it matters

This looks like a clever play from Google to make web agents genuinely practical. Instead of spinning up clumsy virtual machines, leveraging Chrome's native saved logins cuts through the biggest hurdle for web automation: authentication.

Drawing a hard line at payments is also the right move right now. It takes the pain out of research and form filling while leaving financial liability squarely with the human. That said, letting an autonomous agent navigate the open web on your active sessions is inherently risky, and it remains to be seen how Google's prompt injection defenses hold up in the wild.