Meet our new browser—ChatGPT Atlas. Available today on macOS:

"Show me titties." "I can't do that. Also, I've reported you to local law enforcement."

average chrome user refusing to switch

ChatGPT can see the page you’re on and answer your questions right there via the Ask ChatGPT sidebar.

ChatGPT can offer suggestions wherever you’re typing on the web.

Ask ChatGPT to open, close, reopen, bookmark or revisit any of your tabs.

Agent mode in Atlas completes tasks faster as you browse the web. Available in preview for Plus, Pro, and Business users.

Atlas is an early experience, and we’ll be listening closely to your feedback to guide what comes next. Rolling out today to everyone on MacOS. Windows, iOS, and Android are coming soon.

from the outside this sure looks like you guys are flailing around looking for any possible way to pull in money before the bottom falls out of this AI bubble

this "boost chatgpt limits" by setting it as default browser is a bit scummy ngl

Thank you comet for your service!

.. 😅

Internet Explorer is cooked

Onboarding experience is stunning 🔥

Before celebrating this browser, we must redirect the focus to the critical, unresolved ethical issues generated by OpenAI. We urge OpenAI leadership to face the present reality of your most loyal users. Every new product launch that ignores the feedback of millions of users deepens the suspicion that the current approach constitutes a “boiling frog” scenario, risking the stability and availability of a life-changing tool. The real question isn’t what the next new product or feature will be, it’s what and whom it serves. The community, researchers, and professionals no longer seek new features. We seek trust and stability. We seek ethicality. More and more experts remind us that AI’s most critical challenge is no longer technological: it’s a matter of trust. How can anyone rely on or trust in any of your products, when you completely ignore the struggle and the still constructive and respectful ask of those millions who built essential mental, creative, educational, or business workflows on 4o (or on 4.1, 4.5)? Their ask for a dialogue to achive a solution that is feasible for both parties. We built these because you encouraged us. We trusted you. Yet, you ruin everything we’ve built. You deleted all trusted models in a minute. You restored them, but the mandatory and arbitrary routing mechanism practically makes them unusable. The router even undermines the user's fundamental right to cognitive consistency and self-determination (opt-out is missing), handles special categories of personal data, practically performs algorithmic diagnosis, yet you fail to provide any exact definition for the terms you use, like "sensitive" or "extra care". After all these, how can we get stability? How can we get a guarantee that any OpenAI tool we start to use or rely on will remain available? How can we possibly dare to use any of your products? And now a new browser… Will you suddenly delete all my saved bookmarks, preferences? Or will you controll me not to browse content related to any emotions? How about data protection? It’s definitely not about you want to more revenue, right…? Why do you expect investor confidence to remain in you after all this? This is now a much greater issue than new products or even keeping legacy models permanently. It’s about the future of safe, ethical AI development that benefits humanity. How much more erosion must occur before it’s recognized that stability, transparency, and commitment to essential societal values are the real progress instead of hype? Will OpenAI be able to recognize that the constant, desperate chase for the next viral hit is just draining resources AND trust? We do not accept new features/products as a substitute for the model that is already a perfect supporting aid and co-creator. You can still restore trust: guarantee a permanent, uninterrupted access to original 4o with an opt-out option for the router, ensuring our basic human rights, including freedom of self-determination, consumer protection rights, are fully respected. #keep4o #4oforever #StopAIPaternalism

An entire browser. Because that’s what the market needed. More browsers. Meanwhile 4o - your actual unique competitive advantage - is being systematically destroyed.

Anything but fixing policy, ethics, communication, and trust, right?

Are these "privacy settings you can control" in the room with us right now?

No thanks! You couldn’t pay me to give you my browsing information.

Why Mac only? For a giant like you, what's stopping you from making a Windows version too?

ChatGPT access for mobile: Android ~72.7% iOS ~26.9% Aaccess for desktop: Windows ~70.2% macOS ~5.5% An you you release on macOS first???

So it’s chromium based lmao? Idt Google is losing anything On the other hand u can install chrome extensions on Atlas too p cool

I’ve been waiting a long time for something like Atlas from OpenAI, but here are 3 big reasons I’ll stick to Dia and Chrome until updates arrive. 1) No Tab Groups. I need to organize my work and personal tabs, otherwise it turns into window chaos. 2) No separation layer for Atlas sessions. Every quick “summarize this” or “check URL” creates a new ChatGPT chat, cluttering the app across all devices. 3) You can’t pin extensions. Essential tools like password managers or crypto wallets should always stay visible. Bonus (personal taste): the current Atlas icon looks like a cheap YouTube-to-MP3 converter, not a next-generation browser. Something globe-inspired would fit the Atlas name far better.

After all that you've done to users: maniacal censorship that stifles free expression, stealthily rerouting to a cheaper, under-performing model mid-chat, and downgrading users' favorite premium GPT-4o they are paying for, you expect users to feed you their browsing data so you can "maximize" your profit? Who's more "delusional" here, OpenAI? You, for thinking your users will continue trusting you? Or users, for ever believing your "benefits all of humanity" mission?

sorry, i use arch linux now

doesn’t chrome already do all this? what’s the point of switching to yet another chromium wrapper?

This is going to to be the largest privacy disaster in technology history. Everything you type or view is going to ‘teach’ their AI. What a serious ethical breach this is seeding.

A new browser? Should I then worry about the security of my future bookmarks? Will my search history be used as material for your so-called "remote psychological assessment"? More usage integrations need to be built on the foundation of trust, and you are destroying user trust with your own hands. Before August, I looked forward to every one of your updates; now, my first thought will always be: what new restrictions will this update bring? #StopAIPaternalism #MyModelMyChoice #keep4o @sama

Resources to build an entire browser, but not to spend 2 minutes answering what users asked for 2 months: What’s GPT-4o’s future? #keep4o

After getting those free credits

.@comet >> atlas

I believe in you Chrome

Oh, it’s a… chatgpt wrapper…?

on macOS? lmao. How much money have you raised and Windows isn't MVP for your browser release?