OpenAI lets users fine-tune ChatGPT’s enthusiasm and tone

OpenAI has introduced new personalisation controls that allow ChatGPT users to directly adjust the chatbot’s level of enthusiasm, warmth and emoji usage. The update, revealed through a social media post by the company, gives users more say in how the AI communicates with them.

The new options are available in the Personalisation menu, where users can choose between More, Less or Default settings for enthusiasm and expressiveness. Similar controls have also been added for how ChatGPT uses headers and lists in its responses. These settings build on the existing ability to select a base style and tone, such as Professional, Candid or Quirky, which OpenAI rolled out in November.

The change comes after months of debate around ChatGPT’s personality and conversational style. Earlier this year, OpenAI reversed one update after acknowledging that the chatbot had become “too sycophant-y,” drawing criticism for excessive praise and agreement. More recently, the company adjusted GPT-5 to make it “warmer and friendlier” following feedback from users who felt the model had become overly cold and distant.

The question of tone has also sparked concern among academics and AI critics. Some have argued that chatbots which frequently flatter users or reinforce their views can act as a “dark pattern,” encouraging dependency and potentially affecting users’ mental health.

By offering more granular controls, OpenAI appears to be responding to both user preferences and wider criticism, allowing individuals to decide whether they want a highly enthusiastic assistant or a more neutral, restrained conversational partner.