OpenAI Fixes ChatGPT’s Em Dash Issue After User Requests

What happened

OpenAI announced that ChatGPT will now stop using em dashes if users specify this preference in their custom instructions. Previously, the AI often inserted em dashes despite user requests to avoid them. This update was confirmed by OpenAI CEO Sam Altman on November 14, 2025.

Why it matters

• The change improves ChatGPT’s ability to follow user formatting preferences.
• It reduces a common stylistic giveaway that a text was AI-generated.
• The update is part of broader efforts to make ChatGPT more customizable and user-friendly.

Key facts

1. 1. The fix was announced by CEO Sam Altman on November 14, 2025.
2. 2. Users must actively configure custom instructions to avoid em dashes; the default setting still includes them.
3. 3. The update is described by Altman as a "small-but-happy win" for users.