Browsers are turning into agents. Literally.
Opera just introduced Neon, calling it the world’s first “agentic” browser.
AI isn’t just inside the browser anymore—it is the browser.
Neon can fill out forms, book trips, shop online, answer questions, write code, and even create games and websites. All on its own. Even offline. It runs on background AI agents in the cloud. Think of it as an evolved version of the old “Browser Operator” concept—but this time with bigger ambitions. It’s not a tool anymore. It’s a platform.
Opera just opened a waitlist and teased a premium subscription—no price mentioned yet. Meanwhile, Arc is ditching its old browser to go all-in on Dia, a new AI-first interface. Google, OpenAI, and Perplexity are already building their own smart browsers.
So now what?
We’ll soon be talking to our browsers. Telling them what to do, not just what to search. And that raises a bigger issue: who decides what the browser should do for us? These agents will act on our behalf—by learning from us. But if they mess up, who’s responsible?
The future of browsing isn’t just clicking links anymore.
It’s trusting an intelligence to make choices for us—on a web where everything’s connected, and nothing’s neutral.