The AI browser wars
AI companies are building ecosystems and your default browser is the front gate.
Something odd happened in October. Perplexity announced Comet, their new browser, was released to the public. OpenAI unveiled ChatGPT Atlas, complete with a polished but awkwardly scripted demo featuring Sam Altman and three nervous developers. Microsoft quietly rolled out Copilot Mode in Edge, which we only discovered by accident when one of us got a notification email. Three AI-powered browsers, all in the same month. That’s not a coincidence. It’s a land grab.
We’ve seen this movie before. In 2008, Google launched Chrome, and everyone asked, “Why does a search company need a browser?” The answer became clear: control the browser, control the internet. Chrome became the gateway to Google’s ecosystem, and later, the foundation for their hardware expansion. Now, the AI companies are following the same playbook. This isn’t about making browsing better. It’s about capturing you before you even think about using someone else’s AI. Let’s talk about why that matters.
Three browsers. Three strategies.
During our conversation, one of us admitted: “I still default to Google out of habit, because it’s been in my system for years.” That muscle memory is hard to break. But the ‘issue’ with Google is that it works through keywords (and don’t get us started on the promoted and advertorial results you have to fight through first). AI browsers flip that model. Instead of guessing the magic words Google wants to hear, you just ask the question in your own words. It understands what you mean and helps you find it.
All three browsers share the same core premise: AI shouldn’t live in a sidebar. It should be the interface. When you open a new tab in Comet, Copilot Mode, or Atlas, you don’t see Google’s search bar. You see your AI agent, ready to work. This means the question “Should I Google or should I use AI?” slowly disappears. You’re just asking, because that’s the path of least resistance. That’s the pitch, anyway. In practice, each takes a slightly different approach.
Perplexity Comet
Instead of adding AI features to existing browsers, Perplexity’s Comet was designed from the ground up as an AI-first experience. The installation itself signals this: it’s visually striking and polished, with an onboarding flow that feels more like launching a product than installing software. Functionally, Comet positions itself as the research-focused option, built around accuracy and source transparency. It excels at pulling information from verified sources and delivering answers very quickly. It’s the kind of browser you use when you need complete, sourced answers and don’t want to manually cross-reference tabs.
ChatGPT Atlas
The announcement video for ChatGPT Atlas was very interesting. Instead of a smooth polished corporate presentation, Sam Altman sat on a couch with three developers who looked visibly nervous. The vibe was less “polished product launch” and more “let’s gain trust by showing how normal we are, just like you.” But of course the underlying capability matters: Atlas is not just answering questions through ChatGPT, it’s also doing things for you. Agent Mode gives ChatGPT access to a virtual sandbox environment where it can navigate websites, fill out forms, and interact with pages on your behalf.
Copilot Mode in Edge
Copilot Mode in Edge is the enterprise play, leaning hard into organizational integration. Unlike the other two browsers, which treat AI as a personal assistant, Copilot Mode connects directly to your Microsoft 365 environment. It can access all your open tabs simultaneously, not just the active one, which means you can ask it to compare products across multiple windows or summarize information from everything you have open. It integrates directly with Microsoft’s ecosystem, which means it has access to your work data, your SharePoint sites, your Teams conversations. The “Journey” feature tracks your browsing activity and clusters it into task-focused groups with AI-generated summaries and suggested next steps. After a research session, Journey cards appear on your new tab page, nudging you to pick up where you left off. For organizations already embedded in the Microsoft ecosystem, this tight integration is the point: Copilot Mode doesn’t just browse the web, it browses your work. Whether that’s a productivity breakthrough or a governance nightmare we’ll leave in the middle for now.
Why this is really about ecosystems, not features
All three browsers share one goal: they want to be your default. Not your “sometimes” browser. Your one and only. Because once your habit shifts from typing in Google to chatting with an AI agent, you’re locked into their ecosystem. And if tech history teaches us anything, it’s that ecosystems are sticky.
We kept coming back to Google’s Chrome playbook. Chrome didn’t win because it was a better browser; it won because it was fast, free, and instantly glued to everything Google did. Extensions, synced profiles. That seamless integration grew deeper over time. By the time Google shifted to hardware, Chrome was already the glue. The browser was the beachhead.
AI companies follow the same script. Using ChatGPT’s browser? You’re going deep with OpenAI. Pick Perplexity? You’re feeding their model (unless you’re paying to opt out). Opt for Copilot Mode in Edge and Microsoft 365 becomes your new home base. So it’s not the features that matter most. The real question is: which ecosystem do you want to live in?
That’s why these launches came so close together. No player can afford to blink and miss the chance. The browser that hooks users over the next two years might decide how we all interact with AI for the next decade. And unlike the last browser war, there’s no obvious winner yet. This race is real, and truly open this time.
The freemium trap and what you’re actually trading
One of the sharpest tensions we explored was around free versus paid tiers. All three browsers offer freemium models. You can use them without paying. But what are you giving up? With Perplexity and ChatGPT, the trade is data. If you’re not paying, your interactions feed the model. That’s the deal. For personal use, most people are fine with that. You Google your symptoms, you ask ChatGPT for recipe ideas, and you accept that someone, somewhere, is learning from your behavior.
But in a work context? That changes everything. One of us put it bluntly: “I would never use a freemium AI to upload client data. That’s ethically indefensible.” In an enterprise environment, you need guarantees: data stays in your tenant, it’s not used for model training, you can audit access. Microsoft 365 Copilot offers that. OpenAI’s enterprise tier offers that. But the free consumer browser? No chance.
Here’s where the economics get interesting. The companies offering freemium browsers are betting that most users won’t pay. And that’s fine, because those free users generate the training data that makes the paid product better. The more people use the browser, the faster the model improves, the more compelling the paid tier becomes. It’s a flywheel. But it also means the “winner” of the browser war might be whoever can afford to subsidize free users the longest. That’s a capital game, not a product game. And it heavily favors the companies with the deepest pockets: Microsoft, Google, OpenAI (via Microsoft’s investment).
We also debated whether we’re entering a world where different people use different browsers for different contexts. Work in Edge. Personal research in Perplexity. Creative brainstorming in ChatGPT. That fragmentation might be inevitable, but it’s the opposite of what these companies want. They want singular, habitual usage. They want to be your default everywhere.
Is the browser the new operating system?
In the 1990s, the browser wars were about controlling access to the web. Netscape versus Internet Explorer. Microsoft won, then got complacent, and Google swooped in with Chrome. This time, the stakes are higher. The browser isn’t just a window to the web. It’s the interface to AI, the gateway to your data, and the foundation for the next generation of digital work. Whoever wins this war doesn’t just own your default tab. They’re shaping how you work online, and how you interact with information every day. The browser becomes less a passive window, more an active part of your daily flow and decision-making.
Right now, all three browsers are shipping new features as fast as they can, fighting for your attention. But they’re also all betting on the same future: a future where you don’t search, you ask. Where you don’t browse, you collaborate. Where the AI isn’t a tool you open, but the environment you work in. That future feels a lot closer than it did a year ago. And whichever browser gets there first (or gets there best) might define the next decade.
What to watch next: We’re curious whether these browsers will start adding hardware integration. Google went from Chrome to Chromebooks. Will OpenAI build a ChatGPT-native device? Will Microsoft push Windows 11 Copilot Mode as the default experience? And crucially, will privacy regulators step in or will the market decide who wins?
For now, you’ll find us right there, browser open, watching every move.