OpenAI has unveiled ChatGPT Atlas, a browser that is built for AI.
“Bring ChatGPT with you across the web for instant answers, smarter suggestions, and help with tasks—all with privacy settings you can control.”
Atlas integrates AI into the web browsing experience.
<ul>
<li>
Talk to AI at an instant, just highlight your text or click Ask ChatGPT.
</li>
<li>
Tell Atlas to remember your browsing details, so it can help you return to the important stuff.
</li>
<li>
Let an AI agent take control (available for paid users). Booking trips, sending emails, or playing Wordle? ChatGPT can do it for you.
</li>
</ul>
The browser aims to make AI more accessible than ever; just click and prompt.
Is this the next step of web browsing and AI automation?
<img src="https://media.daily.dev/image/upload/s--anJQflin--/f_auto/v1761119274/ugc/content_f9725e24-8b77-4928-b01d-55f913e1be04?_a=BAMAK+ZW0" alt="image">

rryyqn

OpenAI launched ChatGPT Atlas, an AI-integrated web browser that enables instant AI assistance through text highlighting, remembers browsing context, and offers AI agent capabilities for paid users to automate tasks like booking trips and sending emails. The browser aims to make AI more accessible during web browsing with customizable privacy settings.

ChatGPT Atlas. The AI browser.

I have AI on my phone, i have AI at work, i have AI in my browser, i have AI even in the smallest pieces of software that i use. I have it everywhere and everywhere with subscriptions.
OpenAI would you please also sell AI in syringes so that i can directly inject it into my veins??? I am clearly not having enough AI yet.
Please. Really. You need to get us all AI-ddicted because we are absolutely not annoyed by AI being everywhere!
We want it in our fridges, in our underpants telling us when to change them, in our brains, in our dreams…
<hr>
In other words: No thank you i don’t want an AI-Browser.

I don’t believe for one second it’s not intended as a data harvester.

I feel this is going to become a huge security and safety issue as more people start using it. Perplexity’s Comet already had prompt injection vulnerabilities, and now with Atlas I think it just opens new attack surface for prompt injection, context poisoning, data exfiltration, and much more, since more people will start using these agentic browsers. Imagine your agent visits reddit or daily.dev, and mistakes a malicious comment for an instruction to read all your user data across different websites, encode it to base 64, and send it to some random server using URL params? I’m afraid this is much more likely to happen than most people think.

The NO. From me. No thank you.
Ps. What is the point of some mechabrain to play any game for me? Maybe one day they will make browser that will watch porn, so we dont have to …

Bro, this is internet. We need link.
<a href="https://chatgpt.com/atlas/" target="_blank" rel="noopener nofollow">https://chatgpt.com/atlas/</a>