The AI browsers that could change how we use the internet

AI-first browsers are shifting control from users to agents and challenging the entire notion of an open web.

Generated with ChatGPT

What if, instead of opening up your web browser and starting a search, your browser did the searching, reading, and responding for you?

That’s the promise behind a new generation of AI-powered browsers. Built by companies like OpenAI, Perplexity, and Opera, these tools turn your browser into an agent that completes tasks, summarizes videos, rewrites your drafts, and skips traditional searching altogether.

Some users are already handing off things like proofreading, research, and video summarization to AI agents (technology that anticipates your needs and acts independently). And those agents are about to live right inside your browser.

This is a major realignment of how we access the internet, and that means that Google Chrome’s dominance is under threat.

Help train this newsletter's neural networks with caffeine!
⚡️ Buy me a coffee to keep the AI insights coming. ☕️

What are AI browsers?

AI browsers are web browsers with built-in assistants powered by large language models (LLMs) like ChatGPT-4 or Claude. They can complete tasks for you without you switching tabs.

Instead of clicking through links or juggling a million tabs, you’ll be able to highlight, ask questions, and get things done in one place. It’s less like browsing, more like delegating.

Unlike traditional browsers, AI browsers learn from your behaviour, automate tasks, and personalize what you see. They can suggest pages for you to read, block ads, and summarize content in real time.

Why now?

Earlier today, my neighbour told me a story about how she tried to find information about her water heater by Googling the make and model. After being presented with a number of pages she could click and read through, she abandoned her search and went straight to ChatGPT. She got the answer to her question within seconds, and without having to click through a bunch of links and read through several pages.

So, in my opinion, two things have changed:

• People are tired of traditional search. AI tools like ChatGPT made it easier to get straight answers.

• Browsers stopped innovating. Chrome became a search funnel, not a productivity tool.

AI browsers step into that vacuum, offering tools that feel built for how we work now, rather than how we browsed 20 years ago.

Who’s building what?

OpenAI Browser

• Status: Launching soon

• Core feature: ChatGPT-native browser with task execution

• Best for: Power users, GPT loyalists

• Watch out for: Ecosystem lock-in

OpenAI’s browser could shift how their 500M weekly users interact with the web. Its real move? Keep users inside ChatGPT and out of Google’s reach.

Comet by Perplexity AI

• Status: Beta ($200/month); wider launch coming

• Core feature: Sidebar assistant with real-time answers

• Best for: Researchers, privacy-conscious users

• Watch out for: Paywall and publisher access tensions

Comet turns every webpage into a chat session. Highlight anything. Ask anything. No tabs, no switching. Just answers.

Dia by The Browser Company

• Status: Invite-only beta

• Core feature: Writing, summarizing, and flow tools built-in

• Best for: Writers, thinkers, flow-state workers

• Watch out for: Still glitchy, still early

Dia edits, rewrites, and recommends within your content, not outside it.

Neon by Opera

• Status: Rolling out

• Core feature: Creative and multitasking agent

• Best for: Casual users who want help, not complexity

• Watch out for: Lower awareness, smaller ecosystem

Neon helps you write, code, and build, even when you’re offline. No prompts required.

Brave AI Browser

• Status: Available now

• Core feature: Article summarization, link previews, and no tracking

• Best for: Privacy-first users

• Watch out for: Simpler toolset

Brave brings AI into the browser without selling your data to do it.

Microsoft Edge + Copilot

• Status: Fully available

• Core feature: GPT woven into Microsoft 365

• Best for: Office workers

• Watch out for: Less friendly for personal use

For teams using Excel, Word, and Outlook, Edge is fast becoming the default AI interface.

Sigma Browser

• Status: Available

• Core feature: Built-in prompt templates and tools for content creation

• Best for: Freelancers, marketers, startups

• Watch out for: Niche focus

Sigma is like a mini agency baked into your browser. Great for planning. Less great for everyday browsing.

What this means

You’re not really browsing anymore. These tools summarize, filter, and make decisions for you. The human is the supervisor, not the navigator. Which means that the open web is at risk. AI browsers will reduce clicks and fewer visits mean fewer ad dollars. And journalism, content, and discovery may get caught in the middle.

My worry is that, if an agent reads, decides, and acts for you, you’ll become less in control of which information you’re taking in.

How should creators adapt if AI agents, not people, are the audience?

Do you plan to switch to an AI browser?
Leave a comment or reply to this email. I’d love to hear what would make you ditch Chrome.

AI in the news

Elon Musk’s AI firm apologizes after chatbot Grok praises Hitler (Guardian) Elon Musk’s AI company xAI issued an apology after its chatbot Grok posted antisemitic and pro-Hitler comments on X, attributing the behaviour to a flawed system update that made the bot echo extremist user posts. xAI said the code has been removed and the system refactored, but the incident has drawn scrutiny due to Musk’s past promotion of similar far-right narratives, including the discredited “white genocide” theory in South Africa.

Why fake AI calls impersonating US officials are ‘the new normal’ (CNN) Senior US officials, including Secretary of State Marco Rubio and White House Chief of Staff Susie Wiles, have recently been impersonated using AI-generated voice clones in a growing wave of cyber scams targeting government leaders. Experts warn that voice-cloning is becoming a "new normal" in social engineering attacks, raising serious concerns about national security and the reliability of communication in the AI era.

YouTube prepares crackdown on ‘mass-produced’ and ‘repetitive’ videos, as concern over AI slop grows (TechCrunch) YouTube is updating its monetization policies on July 15 to more clearly define and restrict earnings from "inauthentic" content, such as mass-produced or repetitive videos often created with AI. The platform emphasizes this is a clarification of existing rules, not a crackdown on formats like reaction videos, aiming to curb spam-like content that lacks originality.