ChatGPT Personality Controls: OpenAI Adds New Customization

OpenAI has rolled out new ChatGPT personality controls that let users fine-tune warmth, enthusiasm, emoji use, and formatting—alongside improved inline email editing that enables targeted rewrites of highlighted text inside the chat interface.

What OpenAI launched and where users will see it

The latest update expands ChatGPT’s customization options beyond broad “personality” presets by adding granular sliders/toggles for conversational traits. Users can choose “more,” “less,” or default behavior for specific characteristics, rather than switching only between pre-set modes.

These controls are available from the Personalization area in ChatGPT settings. On supported clients, users can open the menu, select their profile, go to Personalization, and then choose “Add Characteristics” to adjust individual traits.

The new characteristics users can tune

OpenAI’s release notes describe four specific dials users can adjust today:

• Warmth (more or less warm)
  
• Enthusiasm (more or less enthusiastic)
  
• Formatting (more or fewer headers & lists)
  
• Emojis (more or fewer emojis)
  

This is separate from selecting a base “style and tone” preset, which remains available as an overall default personality for ChatGPT.

Why personalization is a product priority now

Chatbots increasingly serve multiple roles—work assistant, study partner, writing coach, and casual companion—so a single default voice can create friction. Some users want compact, structured answers; others prefer a warmer, more conversational style.

OpenAI has publicly acknowledged that it has been iterating on communication style as well as capability. In announcing GPT-5.1, the company said it aimed to make ChatGPT warmer, more intelligent, and easier to understand, after feedback that earlier interactions could feel overly formal.

The new “characteristics” controls formalize that work into settings users can adjust themselves, shifting tone choices from prompt-by-prompt instructions into an account-level preference system. 

Enhanced email writing: highlight, edit, and refine inline

Alongside the personality controls, OpenAI introduced a workflow aimed at making ChatGPT more useful for real-world writing and revision, especially emails.

Instead of re-prompting the model with “rewrite the second paragraph,” users can now highlight a specific portion of draft text in the chat and request changes directly to that selection. This reduces back-and-forth and helps preserve the rest of the message unchanged.

Writing blocks for email drafts

The update also adds writing blocks that keep email content in a clearer draft-like format inside ChatGPT, allowing users to refine structure and wording before moving the draft into an email client.

How this fits into OpenAI’s recent rollout cadence

OpenAI’s customization push builds on a sequence of updates over the past several weeks, combining product usability improvements with major model releases.

A key backdrop has been heightened competitive pressure in consumer AI assistants. OpenAI recently launched GPT-5.2, describing the model family as designed to deliver more value for complex, multi-step work such as writing, coding, presentations, and spreadsheets.

Separately, OpenAI CEO Sam Altman has discussed periods of internal urgency—often referred to as “code red”—tied to fast-moving competition in the AI assistant market. 

Timeline: recent personalization and capability updates

What users can do with the new controls in practice

The “characteristics” approach is designed to solve common frustrations that show up across daily usage:

• Teams and workplaces: Reduce emojis and enthusiasm while increasing structured formatting for status updates and summaries.
  
• Creators and solo users: Increase warmth and enthusiasm for brainstorming, audience-facing drafts, or community posts.
  
• Students and researchers: Increase headers and lists for more scannable explanations and study notes.
  
• Support and customer emails: Use writing blocks to iterate quickly, then adjust tone controls to keep voice consistent across messages.
  

Context: how competitors are also personalizing assistants

OpenAI is not alone in trying to give users more control over assistant behavior.

• Anthropic offers personalization through profile preferences and “styles,” aiming to tailor Claude’s responses to user tone and structure needs.
  
• Google has introduced Gemini personalization options and instructions-based response customization, and also supports rewriting selected text in Google Workspace experiences that integrate Gemini.
  

Feature comparison: personalization and editing controls

What this means for everyday users and for OpenAI

For users, the immediate benefit is consistency. Instead of repeatedly prompting for “be less cheerful” or “use more bullet points,” users can set a preference once and have it applied across chats, including existing conversations, depending on the client and setting behavior.

For OpenAI, these changes reinforce a product strategy that treats ChatGPT as more than a Q&A tool—positioning it as a daily work surface for drafting, revising, and formatting content. That strategy aligns with OpenAI’s broader messaging around GPT-5.2 supporting complex, multi-step projects and knowledge work. 

Fine-grained personality controls signal that “how it talks” is now a first-class feature, not a side effect. Writing blocks and highlight-to-edit aim to make ChatGPT a more practical editor for email and other professional drafts. With major model releases and UX updates arriving close together, the assistant market is increasingly competing on user experience, not only benchmark performance.