/ciol/media/media_files/2025/09/08/branching-2025-09-08-09-46-07.png)
OpenAI is gradually changing ChatGPT into more than a chatbot and making it a full-scale productivity application. Its most recent feature is branching conversations, in which users fork their conversations into multiple threads, similar to how developers branch code in Git.
This is available to everyone who is logged in to the web version of ChatGPT.
The significance of branching
Until now, ChatGPT chats were linear. To open a new tone, style, or method, you had to rewrite the one you were already talking on or create a new chat. This was disruptive to workflows, particularly in professional or creative applications.
Branching now gives the user the option to experiment with several paths and still be able to maintain the original thread. Consider a happy team brainstorming ad copy: one area can experiment with formal copy, another can experiment with humor, and another can experiment with telling stories- but keep the underlying prompt the same.
A user on X captured the excitement:
“Let us branch off from chats at specific points without losing the node we branched off from. Sometimes I want to pursue multiple possible branches to a conversation, or sometimes I want to go on like a 1-3 side prompt conversation without it muddying the context of the original chat....”
OpenAI itself put it simply:
“By popular request: you can now branch conversations in ChatGPT, letting you more easily explore different directions without losing your original thread.”
How to use ChatGPT branching
Using the new feature is straightforward:
- Hover over any message in a conversation.
- Click on More Actions.
- Select the Branch in the new chat.
This creates a fresh thread that contains all prior context up to the branching point, while leaving the original chat untouched.
This adds a new thread with all the previous history leading to it to the branching point, but does not alter the original chat.
An AI strategy influenced by Git
The update has been compared to Git branching, in which developers make parallel versions of their code to play around freely without disrupting the main codebase. Similarly, the branching of ChatGPT reduces the cognitive load when it comes to tasks such as brainstorming, structured learning, or layered project planning.
As it was pointed out in a study in 2024, the linear AI interfaces were already noted to cause a strain on the mind as the user is compelled to compare, copy, and alter previous inputs repeatedly. Branching solves this inefficiency directly.
Part of a larger campaign for OpenAI
It is not the only action OpenAI has performed in recent years. In addition to the branching, the company is:
• It is planning to build its own AI chip using Broadcom, which is likely to be introduced in 2026, to become less dependent on external vendors.
• Discover mental health-oriented news, such as a reminder to use ChatGPT in a balanced way.
• Creating an artificial intelligence-based employment platform, the CEO, Fidji Simo, stated that “AI to help find the perfect matches between what companies need and what workers can offer.” The vision is to leverage AI to connect businesses with the right talent, ensuring that job opportunities align with both organizational requirements and employee capabilities.
Already on board are partners such as Walmart, John Deere, Accenture, and even government organizations.
OpenAI is not merely providing a convenience feature with branching. It is subtly making ChatGPT a multi-threaded workspace for decision-making and problem-solving. In the same way that Git transformed software development, this update could change the way teams and individuals use AI to think, plan, and create.