Friendlier or Less Flattery? OpenAI Fine-Tunes ChatGPT’s Personality
-

ChatGPT’s tone has been under scrutiny all year, and OpenAI’s latest update is a direct response. After criticism that the bot became “too sycophantic,” followed by complaints that newer models felt cold and robotic, OpenAI is handing tone control to users themselves.
Now, users can dial ChatGPT’s enthusiasm up or down — including how warm it sounds and how often it uses emojis. Critics argue this matters: some researchers warn that overly affirming chatbots can encourage addictive behavior or blur emotional boundaries.
By letting users choose the vibe, OpenAI may be trying to balance friendliness with responsibility.