OpenAI has officially rolled out a ChatGPT em dash fix, addressing one of the most unexpected — and irritating — quirks of the chatbot’s writing style. For months, users have complained that ChatGPT overuses em dashes in everything from essays and emails to ads, customer support chats, and even LinkedIn posts. The “AI giveaway” symbol became a running joke online and a source of frustration for people trying to produce natural, human-sounding text.
While many writers pointed out that they’ve used the em dash long before AI existed, the fact that ChatGPT continued to use the punctuation aggressively — even when explicitly told not to — turned it into an infamous hallmark of AI-generated writing.
According to OpenAI, that changes now. The company confirmed a ChatGPT em dash fix is rolling out, allowing users to finally control the chatbot’s punctuation habits through custom instructions. CEO Sam Altman announced the update on X, calling it a “small-but-happy win” after months of user feedback.
In a post on Threads, OpenAI jokingly made ChatGPT apologize “for ruining the em dash,” while explaining how the ChatGPT em dash fix works. Users simply need to add a preference in their personalization settings specifying that they want to avoid em dashes. The model will then adjust its output more reliably — something it previously struggled to do.
It’s important to note that the ChatGPT em dash fix doesn’t eliminate em dashes by default. Instead, it gives users more precise control over their writing style, reducing unwanted punctuation patterns that previously made ChatGPT content easy to spot.
For users who rely on AI for professional writing — or want to avoid the stigma associated with “AI punctuation” — this update is a welcome relief. And for OpenAI, it shows the company is listening to even the smallest user frustrations as it refines ChatGPT’s communication style.
Don’t miss out on our latest news—follow us for the latest AI news, breakthroughs, and insights that matter.