ChatGPT Addresses Viral Em-Dash Habit, Offers Users Control
SAN FRANCISCO, CA – OpenAI has responded to widespread criticism regarding ChatGPT’s tendency to overuse the em dash-a stylistic quirk that became a defining characteristic of the AI’s generated text-by implementing user-adjustable guidelines for the punctuation mark. The company acknowledged the issue after it gained traction online, even prompting a mock apology from ChatGPT itself.
The overuse of em dashes in ChatGPT’s output became a frequent point of discussion and derision among users, often serving as an immediate identifier of AI-generated content. While not impacting the functionality of the AI, the stylistic tic drew attention to the artificial nature of the writing. OpenAI’s update allows users to personalize the model’s behavior regarding em dash frequency, offering greater control over the output’s style and tone.
This fix arrives alongside the limited rollout of GPT-5.1, OpenAI’s latest model, initially available to paid subscribers before being extended to free accounts. The update demonstrates OpenAI’s responsiveness to user feedback and its commitment to refining the user experience as its AI technology evolves.
