OpenAI announced in a post on social media that it has launched a new feature that allows users to customize how ChatGPT interacts with them in their conversations. Users can now enjoy an interactive experience that suits their personal preferences, and they can control ChatGPT’s tone and style so it feels friendly or professional, depending on what they want.
The new feature includes multiple personalization options. It lets users adjust ChatGPT’s tone during conversations; for example, they can change the level of warmth and enthusiasm in its replies. Similar adjustments are also available for how ChatGPT uses headings, lists, and emojis in responses.
All of these options are found in the personalization menu. For instance, a user can click on one of the following categories (warmth, enthusiasm, headings and lists, and emojis), then choose one of: Default, More, or Less. Users can also change ChatGPT’s overall style and tone as they prefer, with many options such as professional, friendly, candid, funny, sarcastic, and efficient, some of which the company introduced recently.
Tone and style have been an ongoing issue for ChatGPT throughout this year. OpenAI rolled back one previous update because it seemed “too sycophantic” toward users, then adjusted the GPT-5 version to be more “warm and friendly” after many users complained that the new model felt cold and less friendly in its replies and interactions. Some academics have previously warned that chatbots praising users and complimenting them and their opinions can lead to addictive behavior and may negatively affect users’ mental health.
This feature launch comes as part of OpenAI’s efforts to improve user experience and interaction with artificial intelligence tools. Earlier this December, OpenAI signed a strategic partnership with Disney, which granted the company access to Disney characters and content such as Simba, Mickey Mouse, and other iconic characters. The agreement allows Sora users to create content using more than 200 different Disney characters.