OpenAI just solved one of ChatGPT's most annoying quirks. The company announced that users can now personalize their ChatGPT settings to stop the AI from stuffing em dashes into everything it writes. It's a small fix that addresses a surprisingly big problem that's been driving users crazy for months.
OpenAI just delivered what CEO Sam Altman calls a 'small-but-happy win' - ChatGPT will finally stop bombarding users with em dashes when they ask it not to. The announcement comes after months of frustrated users trying unsuccessfully to get the AI chatbot to dial back its obsession with the punctuation mark that became known as the 'ChatGPT hyphen.' In a post on X, Altman confirmed that telling ChatGPT not to use em dashes in custom instructions now actually works as intended. The em dash epidemic had reached a tipping point across the internet. Rolling Stone and The New York Times both documented how the telltale punctuation mark started appearing everywhere - school papers, work emails, customer service chats, LinkedIn posts, and online forums. People began using em dash frequency as an unofficial detector for AI-generated content. The problem wasn't just aesthetic annoyance. Writers found themselves accused of laziness and AI reliance simply because their natural writing style included em dashes - a punctuation mark that existed long before large language models adopted it. But ChatGPT's inability to stop using the symbol, even when explicitly instructed not to, made any em dash suspect in the eyes of readers. OpenAI's community forums filled with complaints from users who couldn't get ChatGPT to follow their formatting preferences. Multiple threads documented the chatbot's stubborn refusal to comply with anti-em-dash instructions, suggesting a deeper technical issue with how the model processed style guidance. The fix represents more than just punctuation control - it signals progress in AI personalization and instruction-following capabilities. OpenAI explained in a Threads post that ChatGPT will now better respect custom instructions about em dash usage through personalization settings. The company even had ChatGPT apologize for 'ruining the em dash' in what appears to be a tongue-in-cheek acknowledgment of the controversy. However, the solution isn't automatic. Users must actively configure their custom instructions to avoid em dashes - the default behavior hasn't changed. This means new users will still encounter the punctuation mark unless they specifically opt out through their settings. The timing of this fix aligns with broader push toward more customizable AI experiences. Recent updates have focused on giving users greater control over ChatGPT's tone, style, and output format. The em dash solution fits into this pattern of responding to user feedback about AI behavior that feels too rigid or predictable. Industry observers see this as part of a larger trend toward AI systems that can better mimic individual writing preferences rather than defaulting to a standardized voice. The em dash controversy highlighted how small stylistic choices can significantly impact user experience and trust in AI-generated content. For content creators and professionals who rely on ChatGPT for drafting, the ability to control punctuation preferences could help maintain their authentic voice while leveraging AI assistance. This matters especially in professional contexts where writing style contributes to personal branding and credibility.
