
AI Browsers: Your New Digital Assistant at Work
Think about how much of your day happens inside a browser. Reading news. Managing email. Shopping. Planning trips. Researching for work. The browser has always been the window to the internet—but until now, it hasn’t done much for you.
That’s changing. A new wave of AI browsers is emerging, designed not just to display the web but to help you navigate and act on it. They’re called “agentic browsers” because they behave like agents: able to summarize, compare, fill forms, and in some cases, even complete tasks on your behalf.
It’s like going from driving a stick shift to sitting in a Tesla on autopilot. You still steer when needed. They are not at their full potential yet, but I think adopting them now instead of waiting for them to become a serious player will put you ahead.
What exactly can an AI browser do?
Here are a few real examples of how AI browsers are being used today:
One of the most important things these browsers do is make querying effortless, much like Google Chrome did with Google search. In Perplexity Comet, for instance, there’s an “Assistant” button. Clicking it opens a right-side pane where you can immediately start interacting with Perplexity. Similarly, in Dia, opening a new tab presents you with a large search box where you can type your prompt. Once you begin, the full power of an LLM is right at your fingertips.
At first glance, this may feel no different than opening chatgpt.com and typing a prompt there. But in practice, it’s easier. You no longer need to take that extra step of typing in a website address—it’s all integrated directly into your browsing experience.
The second big advantage is automation. I’ve used Perplexity Comet to handle tasks like archiving my emails, pulling together all communications from a specific person over the last three months, and even accepting LinkedIn invitations. These browsers can actually do things for you. While not every feature works perfectly yet—and actions can feel a bit slow—the concept itself is exciting. The shortcomings will likely disappear as the technology matures, making these tools faster and more reliable.
Here are a few more things you can do with them:
- Summarize quickly: Instead of reading through a 2,000-word article, you can ask the browser to give you the three key takeaways.
- Cross-tab comparisons: Open a few product pages, then simply ask, “Which one has the best reviews under $500?” The AI scans all tabs and answers instantly.
- Bookings and shopping: Say, “Book me a flight Friday evening to San Francisco, return Sunday night, under $300.” The browser searches, compares, and starts filling in forms. You confirm before checkout.
It’s not magic, but it’s close to having a digital assistant sitting beside you.
The main players to watch
Several companies are racing to define this new category. Each has a different flavor:
- Perplexity’s Comet
The most “hands-on” agent so far. It can actually do multi-step tasks like filling out checkout forms or drafting email replies based on your inbox. It’s aimed at power users, but comes at a premium: Comet access today is tied to Perplexity’s $20/month Pro plan. I use Comet as my default browser. - Opera Neon
Opera calls it their “agentic browser” and it’s built around three modes: Chat (answer questions), Do (act on sites—like buying or booking), and Make (create things in the cloud, like reports or prototypes). It’s still in alpha with a waitlist, but it points toward the boldest vision of automation. - Dia (from The Browser Company)
If you know Arc, think of Dia as its AI-native successor. And if you don’t know Arc, don’t worry, I am sure you know Atlassian! Atlassian recently bought The Browser Company. It’s not as autonomous as Comet or Neon, but it’s wonderful as a thought partner. You can “chat with your tabs,” use Skills for repeatable workflows, and get writing or learning help right in the page. It’s free in beta and focused more on organization and creativity than automation. - Microsoft Edge (Copilot Mode)
Built into the mainstream Edge browser, this is the easiest entry point: free, available today, and steadily gaining agent-like features. It can summarize across tabs, handle voice commands, and is testing light action-taking like bookings. It’s less specialized, but perfect if you just want a taste without switching browsers.
My take: which one do I like?
I’ve experimented with both Dia and Perplexity Comet, and so far I prefer Perplexity Comet. Dia’s approach to automation is slightly different—you have to build what it calls “skills” and then invoke them with specific commands. For example:
/write Draft a LinkedIn post about the benefits of AI browsers for property managers.
Here, /write is the command (or skill), and the rest of the prompt follows after it. Personally, I find this inconvenient because it requires remembering exact commands.
With Perplexity Comet, there are no such commands. You simply interact with it as you would with ChatGPT or any other LLM. This makes the experience more natural and intuitive. Beyond that, Comet can actually take actions on your behalf—like sending an email—something Dia hasn’t yet learned to do.
For that reason, at this moment, Perplexity Comet is my favorite.
The bottom line
AI browsers are the next step in how we interact with the internet. For the first time, your browser isn’t just a passive window—it’s an active partner. Whether you want faster summaries, smarter comparisons, or actual task automation, these tools are starting to deliver.
For professionals – the group balancing career demands, family logistics, and constant information flow—this shift matters. An AI browser won’t replace your judgment, but it can save you time, reduce mental clutter, and free up energy for the things that matter most.
The era of “agentic browsing” has begun. The question is: how soon do you want your browser to start working with you?
Leave a Reply