ChatGPT Users Can Now Remove Em Dash's Humbature Signs From AI-Made Text

JAKARTA OpenAI says that ChatGPT's habit of including long connection signs or em dashs has been fixed. Now, users will not see their readings if they ask to be removed.

Em dash is often associated with artificial intelligence (AI) generative content because these readings often appear in conversational captions with chatbots. Many don't like these punctuations because their content looks like AI-made.

This problem also cannot be solved by ChatGPT users, unless they edit the text manually again. The reason is, even though the user asks ChatGPT not to use the reading sign, the em dash still appears between the texts.

Fortunately, OpenAI listens to the complaints of its users. On Friday, November 14, 2025, OpenAI CEO Sam Altman said that the company managed to overcome the problem. This is good news for loyal ChatGPT users.

"If you tell ChatGPT not to use em dash link marks in your special instructions, eventually ChatGPT will do what it should," Altman said in a post on X.

It is important to know that the em dash will not be automatically deleted. At least, users can avoid the presence of em dash if the user gives instructions via a special instruction setting.