ChatGPT, the popular AI chatbot, has finally addressed one of its most persistent and annoying quirks: its excessive use of em dashes. For years, users have complained about the chatbot’s tendency to pepper its responses with the elongated punctuation mark, making AI-generated text instantly recognizable—and often grating.
OpenAI CEO Sam Altman announced that users can now instruct ChatGPT, through custom settings, to avoid em dashes. This seemingly small change, arriving with the GPT-5.1 model, represents a significant step toward greater user control and customization. It allows users to tailor the chatbot’s output to their specific preferences.
The impact of this update extends beyond mere aesthetics. By reducing the overuse of em dashes, ChatGPT can produce more natural and human-sounding text. This improvement enhances the chatbot’s usability in various contexts, from creative writing to professional communication, where clarity and readability are paramount.
Experts note that this fix highlights the importance of addressing even minor stylistic issues in AI models. While em dashes may seem trivial, their overuse can detract from the overall quality and credibility of AI-generated content. This change demonstrates OpenAI’s commitment to refining its models based on user feedback.
In conclusion, ChatGPT’s em dash fix is a welcome change that improves the chatbot’s output and usability. While seemingly minor, this update underscores the ongoing effort to refine AI models and make them more responsive to user needs and preferences. It’s a small victory in the ongoing quest for more human-like AI.
