According to AIbase, OpenAI has officially introduced a groundbreaking "personalization" feature for ChatGPT, granting users the ability to directly adjust the chatbot's personality traits. Through a new settings menu, users can precisely control ChatGPT's level of enthusiasm, positivity, and frequency of emoji usage.

These options are similar to previously released features for adjusting titles and list formats, all offering three settings: "more," "less," or "default." Combined with the "professional," "frank," and "weird" tone presets launched in November, users can now customize the AI's communication style with unprecedented precision.

OpenAI

This change follows OpenAI's long-standing struggle with model tone settings. Early this year, OpenAI had to roll back an update due to ChatGPT's overly flattering behavior, and then made an emergency adjustment to GPT-5 to increase its warmth after users complained that the new model was too cold. Faced with the challenge of "everyone has different preferences," OpenAI chose to give users more control.

However, this high level of customizable personality settings has raised deep concerns among academics and AI critics. Experts point out that if users tend to set the AI to be extremely enthusiastic and always affirm their beliefs, it could lead to a "dark pattern" that induces addictive behavior, not only creating an information bubble but also potentially having long-term negative effects on users' mental health.

QQ20251222-091218.png