ChatGPT Finally Lets Users Ban Those Pesky Long Dashes

Image source note: The image is AI-generated, and the image licensing service is Midjourney

In a move celebrated by writers everywhere, OpenAI has given ChatGPT users the power to banish em dashes—those long horizontal lines that had become the AI's linguistic fingerprint. "This finally makes it do what it should," tweeted CEO Sam Altman, calling it "a small but happy victory" for frustrated users.

The Dash Dilemma

The custom instructions feature—buried in ChatGPT's settings—now includes an option to avoid what one blogger described as "the dash plague." For months, users tried everything from polite requests to stern commands without success. "I'd pay $100 to learn how to stop ChatGPT from using em dashes," lamented author Chip Huyen on social media. "It's driving me crazy."

Why did ChatGPT develop this dash obsession? Syracuse University journalism professor Aileen Gallagher explains: "If an AI primarily trains on magazine or blog writing, both styles heavily favor em dashes." These versatile punctuation marks help with rhythm and transitions while being computationally efficient—two qualities AI models naturally prioritize.

Identity Crisis

The dash overload created unexpected problems beyond stylistic preferences. Some writers found their own work under suspicion. "Now everyone assumes I used ChatGPT whenever I include dashes," complained one user who favored the punctuation mark long before AI writing tools existed.

Social media brimmed with memes mocking ChatGPT's dash dependency—until Altman dropped his surprise announcement just one day after releasing GPT5.1. The new model brings faster processing for simple tasks and expanded personalization options alongside its dash-disabling capability.

What Changed?

The update represents more than just punctuation preferences. It signals OpenAI's growing attention to user control over AI writing styles. While dashes serve legitimate purposes—connecting thoughts efficiently or creating dramatic pauses—their robotic overuse had become a distraction.

For professionals who need their writing to sound authentically human, this small setting change makes a big difference. As one relieved user posted: "Finally, I won't have to manually remove hundreds of dashes before submitting reports!"

Key Points:

• User control: New setting lets you disable ChatGPT's signature em dashes
• Style matters: Overuse had become a telltale sign of AI-generated text
• Bigger picture: Part of GPT5.1's expanded personalization features