OpenAI just made ChatGPT better at browsing the web. For anyone with 50+ tabs open, this one's for you.
The Chrome extension allows Side Chat to reference YouTube videos and open tabs. You can highlight text on any page and ask questions.
The desktop app adds URL suggestions as you type, access to your browser history, and granular controls.
All available today.
The feature is useful. But the real story isn't about summarization.
It's about context. ChatGPT can now see where you are, not just what you type.
For tab hoarders, this is a productivity win. For OpenAI, it's a strategic move. The Chrome extension is the entry point. The desktop app is the real play.
OpenAI is embedding itself where you work — not just in a chat window, but in your browsing session.
Privacy controls are included. You choose what it sees. The question is whether anyone reads them.
The AI now has persistent context across your session. The implications go beyond summarization.
ChatGPT is no longer a destination. It's becoming a layer.
The browser is the last frontier for AI assistants. It's where people spend most of their digital lives. It's where context lives — open tabs, browsing history, highlighted text, YouTube videos, documents. All the information you're actively working with.
OpenAI wants to be there. Not as a separate window. As part of the workflow.
This is why the desktop app matters more than the extension. The extension gives you a taste. The desktop app gives OpenAI access to your browsing patterns — with your permission, of course.
The privacy controls are robust. You grant permissions per site. You can allow once or allow always. Browser history access requires explicit consent. But the tension is obvious: the more you give, the more useful it becomes.
The product logic is clean. The user trade-off is real.
Google has been trying to make Assistant work across tabs for years. Microsoft has Copilot in Edge. Both are playing the same game — embedding AI into the browser.
OpenAI just made its move. It's not the first. But it's arguably the most native.
ChatGPT is better than both at reasoning. If it can also see what you're looking at, the combination is powerful.
P.S. Most people will use this to summarize YouTube videos. Power users will use it to build persistent context across sessions. OpenAI knows which group matters more.
