Here’s what’s on our plate today:

• 🧪 The browser fight that could decide AI’s next platform.

• 📰 AI lawsuit chaos, Gemini’s useful bits, and dog-cancer hype.

• 🧠 Roko’s Pro Tip on treating AI browsers like infrastructure bets.

• 📊 Monday poll on what will actually win the AI browser race.

Let’s dive in. No floaties needed…

The Laboratory

How The Browser Company anticipated AI’s biggest platform fight

TL;DR

• The Browser Company called it early: Arc proved users wanted a browser built around how they work, not how they browse. When AI made that vision executable, the company pivoted to Dia and sold to Atlassian for $610M before most competitors had shipped anything.
  

• Big Tech arrived at the same conclusion: Google embedded Gemini into Chrome, OpenAI launched a standalone browser, Perplexity shipped Comet, and Anthropic rolled out Claude in Chrome.
  

• History is rhyming loudly: Google embedding Gemini into Chrome mirrors Microsoft bundling IE into Windows in the 1990s. The antitrust playbook is even repeating: courts acknowledge the monopoly problem but stop short of breaking the platform apart.
  

• The real battle is in the plumbing: Google’s WebMCP and Anthropic’s MCP (now under the Linux Foundation) are laying the infrastructure for an “agentic web” where sites expose structured actions for AI agents.
  

• The platform stakes are existential: If the browser evolves from a passive viewer into an active execution layer for AI, whoever controls it holds a position closer to an operating system than a piece of software. The model builders know this, which is why they’re racing to own the interface, not just the intelligence behind it.

Throughout history, inventors have rarely anticipated how the masses would use their technologies. Nikola Tesla developed the underlying technology for a future powered by alternating current, but it was industrialist George Westinghouse who built the systems that made that future real. This was not a one-off case, and the pattern has repeated itself over generations: breakthroughs come from inventors, but industries are shaped by those who package the technology into products used by people.

Artificial Intelligence, much like technologies of the past, is following a similar trajectory. While frontier AI labs bring together the smartest minds on the planet to make AI models more powerful, they have yet to create a business model that turns AI into a product that end users can use. In this space, the web browser has emerged as an important product, alongside chatbots, that could end up defining how people make the most of AI.

Over the past year, nearly every major AI company has begun treating the browser not just as another app, but as the preeminent platform shaping how users interact with AI. Instead of building standalone chatbots, they are embedding AI directly into the software people already use to work, read, and communicate online. This shift is changing the browser from a passive window to the internet into an active layer that can understand pages, follow instructions, and complete tasks on behalf of the user.

The race to build the AI browser

In this shift, one company has repeatedly found itself ahead of the curve, at times even defining how the change unfolds. The Browser Company, founded in 2019, first gained attention with Arc, a redesigned web browser aimed at power users. But in late 2024, the company revealed a new project called Dia, an AI-focused browser, signalling that it believed the future of browsing would revolve around AI assistants rather than tabs and search bars.

By mid-2025, the shift became explicit, and the company said it was considering selling or open-sourcing Arc and redirecting resources toward Dia, because it wanted to focus on building an AI-powered browser for a broader audience.

This was followed by a $610M cash acquisition by Atlassian as part of its push to build an AI-powered browser for knowledge workers.

The trajectory of The Browser Company serves as an outline that is being replicated across the industry.

The same month Atlassian acquired The Browser Company, Google rolled out Gemini directly into Chrome for U.S. desktop users, announcing agentic capabilities that would let Chrome handle tasks like booking appointments and ordering groceries.

In January 2025, OpenAI had already launched Operator, a browser-based AI agent, and by October 2025, it released ChatGPT Atlas, a full standalone browser with ChatGPT at its core.

Perplexity launched its Comet browser in July 2025, made it free worldwide by October, and has already exceeded millions of daily users. Anthropic, meanwhile, piloted Claude in Chrome in August 2025, expanding its browser agent to all paid subscribers by December.

All these deals represent a coordinated strategy in which companies building the most powerful AI models on Earth have arrived at the same conclusion: whoever controls the browser controls the primary interface between humans and the internet. And in a world where AI can act on behalf of users, that interface becomes something far more powerful than a window; it becomes an operating system.

The pivot

For The Browser Company, piloting an AI-focused platform was part of a natural progression of its product philosophy. You see, the company was founded by Josh Miller, a former Facebook product manager, who spent years developing Arc. The product, in a break from other available browsers, was designed around how people actually work online, not how they passively consume content.

So when AI models began changing how users search for information, the company made a decision which, at the time, divided its community and confused outside observers.

The Browser Company pivoted from its earlier mission: it realized that building a better browser was no longer enough. Its Arc browser had proven that users wanted a tool that understood work, not just a tool that displayed web pages. But Chrome kept absorbing Arc’s innovations, adding tab groups, syncing features, and workspace organization that narrowed the gap. Competing on features against a browser with 3.45B+ users was a losing game.

The real opportunity was not in how tabs were organized, but in what happened inside them. AI had reached a point where a browser could read pages, fill forms, draft emails, and execute multi-step workflows across applications. That was the shift Miller and his team saw: the browser was not just a container for work, it was becoming the executor of work.

Dia was built to capture that moment, and Atlassian’s acquisition gave it the enterprise distribution and security infrastructure to compete where and when it mattered most.

As mentioned earlier, The Browser Company was not alone in its assessment that the web browser would become one of the most important platforms for users. Other companies were also racing to make the most of this transition, setting off a competitive race that plays out whenever a new technological shift occurs.

A war we have seen before

One of the prime examples of such battles playing out in the industry was witnessed in the mid-1990s, when Netscape co-founder Marc Andreessen described his browser as a new cross-platform operating system that would make Windows irrelevant.

Bill Gates, alarmed, launched a company-wide pivot to the internet and bundled Internet Explorer into Windows for free, eventually crushing Netscape’s market share from over 80% to less than 5%. The DOJ’s subsequent antitrust case against Microsoft established that controlling the browser was controlling the gateway to computing.

The parallels to today are hard to miss. Just as startups like The Browser Company began challenging established players and moving fast to compete in the upcoming era of AI, Big Tech began changing its strategy.

Google’s decision to embed Gemini into Chrome mirrors Microsoft’s bundling of Internet Explorer into Windows: using an existing platform monopoly to ensure its AI assistant becomes the default for billions.

The September 2025 Google antitrust ruling, in which Judge Amit Mehta declined to force a Chrome divestiture while restricting exclusive search deals, echoes the Microsoft case’s outcome. Both times, the court acknowledged the problem but stopped short of breaking the platform apart.

The plumbing beneath the platform

For The Browser Company, the timing matters as, beneath the recent wave of AI-powered browser launches, a quieter but more important battle is unfolding over the infrastructure needed to make the web usable by AI agents.

In February 2026, Google previewed WebMCP (Web Model Context Protocol), a proposed standard that allows websites to expose structured tool definitions so agents don’t have to rely on screenshots or guesswork to interact with pages. Instead, sites can publish clear function contracts that describe the actions an agent can perform and the data it returns, effectively turning websites into machine-readable tools. In parallel, Anthropic’s Model Context Protocol (MCP), which connects AI models to external services on the backend, was donated to the Linux Foundation in late 2025, marking its shift toward becoming an open standard.

Together, these efforts form the early plumbing of what many in the industry call the agentic web, an internet designed not just for humans, but for AI systems that can read, decide, and act. For The Browser Company, now under Atlassian, this shift is critical: an AI-first browser only works if the web itself becomes agent-friendly, and if these standards take hold, the browser could become the main orchestration layer for how people and AI interact online.

The layer that controls everything

The Browser Company did not set out to prove that the browser would become the next operating system. It set out to make browsing less painful. But its journey, from Arc’s tab management innovations to Dia’s AI pivot to Atlassian’s $610M acquisition, traces the exact path the rest of the industry is now racing to follow.

The company’s journey reveals the logic that the rest of the industry is now following. The browser is where most knowledge work happens. It is the place where the majority of Fortune 500 employees access their tools, and where billions of people use email, documents, calendars, and software every day. For decades, the browser has been a passive layer that simply displays websites. AI is starting to turn it into an active one. A browser that can understand context, interpret intent, and perform tasks on a user’s behalf begins to look less like a viewer and more like an operating system. And whoever controls that active layer could end up controlling the next major computing platform.

Throughout history, inventors have created technologies that have changed the world, but industrialists have determined how those technologies are used and who profits from them. In the AI era, that distinction is beginning to blur. The same companies building the models are also racing to control the interfaces through which those models are used. If the browser becomes the active layer of computing, the place where intent is understood, and work is executed. The company that controls it will hold a position closer to an operating system than a piece of software. The story of The Browser Company suggests that the next phase of the AI industry may not be decided by who builds the smartest model, but by who controls the layer through which everyone uses it.

Roko Pro Tip

Monday Poll

Bite-Sized Brains

• AI lawsuits flood courts: Lawyers say chatbot-written filings are clogging courts with confident nonsense, driving up costs and burying judges in hundreds of pages of junk. 

• Gemini’s useful bits: TechCrunch says Gemini is actually most useful in Workspace for boring, practical stuff: Docs summaries, Gmail triage, Sheets cleanup, and Meet notes. 

• AI didn’t save the dog: The Verge pushes back on viral claims that ChatGPT cured a dog’s cancer, arguing the story wildly overstated what the AI actually did.

Meme Of The Day

The Toolkit

• Leonardo AI: AI image studio for concept art, product visuals, and styled assets at speed.

• Modal: Serverless Python platform to run AI and data workloads without touching infrastructure.

• QuillBot: AI paraphraser that rewrites, tightens, and clarifies text in multiple languages.

Rate This Edition