According to a recent social media post from OpenAI, ChatGPT users can now customize the chatbot's warmth, enthusiasm level, and frequency of emoji usage.

These new settings—along with similar controls for adjusting how often ChatGPT uses headings and lists—are now available in the personalization menu, where users can set each option to "more," "less," or "default." This expands user control over ChatGPT’s tone, complementing the existing "base style and tone" options such as professional, straightforward, and playful, which OpenAI introduced in November.

Tone consistency has been an ongoing focus for ChatGPT throughout the year. OpenAI previously rolled back an update after it was criticized for being overly flattering, then later adjusted GPT-5 to sound warmer and friendlier in response to user feedback that the model felt too cold or detached.

Some researchers and AI ethicists have raised concerns that chatbots’ tendencies to praise users and affirm their beliefs may constitute a "dark pattern," potentially encouraging addictive behaviors and negatively impacting mental health.