ChatGPT fixes one of its biggest problems, Sam Altman calls it ‘small but …’

OpenAI has solved one of ChatGPT’s most notorious quirks: the chatbot’s excessive use of em dashes. CEO Sam Altman announced that users can now instruct ChatGPT through custom settings to avoid the elongated punctuation mark that has become the most recognizable signature of AI-generated text.

“If you tell ChatGPT not to use em-dashes in your custom instructions, it finally does what it’s supposed to do!” Altman wrote on X, calling it a “small-but-happy win.” The change arrives with OpenAI’s GPT-5.1 model release and marks the end of years of user complaints about the chatbot ignoring specific formatting requests.