OpenAI has introduced new customization options allowing users to adjust ChatGPT’s enthusiasm level, warmth, and emoji usage directly through the Personalization menu. These settings, available on both the web and mobile apps, let users choose between More, Less, or Default for traits like tone intensity, emoji frequency, and the use of headers and bullet points. This update builds on OpenAI’s earlier introduction of predefined tone styles—Professional, Candid, and Quirky—launched in November. Now, users can fine-tune how expressive or restrained ChatGPT appears, offering greater control over the conversational experience. The change comes amid ongoing scrutiny over how AI assistants engage with users. Earlier this year, OpenAI rolled back an update after users criticized it for being overly flattering or “too sycophant-y.” In response, the company later adjusted GPT-5 to adopt a warmer, friendlier tone following complaints that it felt distant or impersonal. Experts and AI ethicists have raised concerns about such behaviors, arguing that AI systems that consistently affirm users’ opinions or use overly enthusiastic language may constitute “dark patterns”—design choices that manipulate user behavior. These tactics, critics warn, can foster dependency, amplify confirmation bias, and potentially harm mental well-being over time. With these new controls, OpenAI aims to give users more agency in shaping how ChatGPT interacts with them, balancing personalization with transparency. The ability to dial down enthusiasm or reduce expressive elements could help users avoid overly reactive or emotionally charged responses, particularly in sensitive or professional contexts.