Anyone who’s used AI for writing has probably noticed one thing: ChatGPT loves its em dashes. So much so that some readers began spotting these dashes as a telltale sign of AI-generated text. But that might change soon, OpenAI has announced that ChatGPT will now avoid using em dashes when users specifically ask it not to.
The em dash has been popular even before generative AI emerged, but its frequent presence in school assignments, emails, comments, customer service chats, LinkedIn posts, online forums, advertisements, and more has turned it into a recognizable indicator of AI-generated content.
The em dash saga became more frustrating when ChatGPT kept inserting them, even when users explicitly asked it to stop. But that’s now history. OpenAI CEO Sam Altman confirmed in a post on X that the issue has been fixed. In a post on X, Altman writes, “If you tell ChatGPT not to use em-dashes in your custom instructions, it finally does what it’s supposed to do,” calling the update a “small-but-happy win”.
ChatGPT’s em dash habit has been a talking point for many users. In a Threads post, the company acknowledged the issue, jokingly apologizing for “ruining the em dash,” and explained that the chatbot can now better avoid using them. While this doesn’t remove em dashes entirely from every response, it gives you more control over how often they appear.
Step-by-Step Guide:
Open ChatGPT and go to Settings.
Navigate to Personalization, then Custom Instructions.
In the section where you can specify how ChatGPT should respond, add a note asking it to avoid or minimize em dash usage.
Save your instructions and start your conversation — ChatGPT will now follow your preference.
Adjust or refine the instructions anytime for better control over punctuation habits.
Meanwhile, OpenAI has launched a pilot for ChatGPT group chats, letting users collaborate in a single conversation. Available since November 13 for all accounts, it’s currently being tested in Japan, New Zealand, South Korea, and Taiwan.