OpenAI Rolls Out Personality Customisation for ChatGPT
OpenAI has launched a significant update to ChatGPT that allows users to assign personality traits and customise the AI's conversational style. The new feature enables users to specify nicknames, professional contexts, and personality characteristics ranging from "Chatty" to "Analytical."
This marks a shift towards more personalised AI interactions, though it operates through prompt engineering rather than fundamental changes to the underlying model. The feature is currently rolling out on ChatGPT.com and Windows desktop, with mobile and MacOS versions following soon.
How Custom Traits Transform AI Conversations
The new customisation system lets users define three key aspects of their ChatGPT experience. Users can assign a preferred nickname for the AI to use in conversations, specify their profession for more contextually relevant responses, and select from personality traits including:
- Chatty: More conversational and engaging tone
- Encouraging: Supportive and motivational responses
- Gen Z: Informal, youth-oriented communication style
- Sceptical: Critical and questioning approach to topics
- Analytical: Logical, data-driven discussion style
- Creative: Enhanced brainstorming and ideation capabilities
- Concise: Brief, to-the-point responses
- Empathetic: Sensitive handling of personal topics
Importantly, this feature operates separately from ChatGPT's memory system. While the AI can remember conversations from months ago, custom traits modify responses in real-time without affecting long-term memory retention.
"We're making it easier for users to have the kind of conversation they want with ChatGPT, whether that's professional, casual, or somewhere in between," said an OpenAI spokesperson in the feature announcement.
By The Numbers
- Eight distinct personality traits available at launch
- Rolling out across ChatGPT.com and Windows desktop first
- Mobile and MacOS versions expected within weeks
- EU users face delayed access due to regulatory considerations
- Feature operates independently of ChatGPT's existing memory system
Technical Reality: Prompt Engineering, Not Model Changes
Despite the marketing around personalisation, this update doesn't fundamentally alter how ChatGPT's large language model operates. The feature relies on sophisticated prompt engineering, essentially providing the AI with detailed instructions about how to respond based on user preferences.
This approach differs from more substantial AI developments like agentic AI systems, which involve genuine changes to how AI models process and respond to information. The custom traits function more like preset conversation starters than deep personality modifications.
"It's a user-friendly wrapper around what advanced users have been doing manually for months," explains Dr. Sarah Chen, AI researcher at Singapore National University. "The real value is in accessibility, not technological breakthrough."
Users seeking more comprehensive customisation might benefit from learning advanced techniques, such as understanding why strategic prompting can make ChatGPT think more effectively.
| Feature Type | Custom Traits | Memory System | Advanced Prompting |
|---|---|---|---|
| Persistence | Session-only | Cross-conversation | Manual setup |
| Complexity | Preset options | Learns over time | Fully customisable |
| User Control | Simple selection | Automatic | Expert level |
Moderation and Safety Considerations
OpenAI has implemented moderation systems to prevent misuse of the personality customisation feature. The company reviews trait combinations to ensure they comply with its terms of service and don't enable harmful or misleading interactions.
This moderation approach reflects broader industry concerns about AI safety, particularly relevant given ongoing discussions about responsible AI governance across Asia. The company aims to balance user customisation with safety requirements.
The feature's rollout excludes EU users initially, likely due to regulatory compliance considerations under the AI Act and GDPR frameworks. This geographical restriction highlights the complex regulatory landscape facing AI companies globally.
Market Impact and User Applications
Early user feedback suggests the feature appeals particularly to content creators, business professionals, and casual users seeking more engaging AI interactions. The customisation options could prove valuable for specific use cases where tone and style significantly impact effectiveness.
However, some users question whether the feature addresses ChatGPT's tendency towards overly deferential responses, a behaviour pattern that has frustrated many users seeking more direct, confident AI assistance.
The update arrives as competition intensifies in the conversational AI space, with companies racing to differentiate their offerings through personalisation and user experience improvements.
How do custom traits differ from ChatGPT's memory feature?
Custom traits modify response style in real-time during conversations, while memory retains information across separate chat sessions. Traits focus on how ChatGPT communicates, whereas memory affects what it remembers about you.
Can I use multiple personality traits simultaneously?
Users can select combinations of traits, though OpenAI moderates certain combinations to prevent conflicts or inappropriate uses. The system is designed to blend compatible traits naturally.
Will this feature affect ChatGPT's accuracy or factual responses?
Personality traits primarily influence tone and style rather than factual content. However, traits like "Sceptical" or "Encouraging" could subtly influence how information is presented or framed.
When will mobile users get access to custom traits?
OpenAI plans to roll out mobile and MacOS support within the next few weeks, following the current web and Windows desktop launch phase.
Are there any restrictions on trait customisation?
Yes, OpenAI moderates trait usage to prevent harmful applications and ensure compliance with terms of service. Some geographic restrictions also apply, particularly for EU users.
The custom traits feature signals OpenAI's recognition that one-size-fits-all AI interactions have limitations. As AI assistants become more integrated into daily workflows, personalisation will likely become increasingly important for user satisfaction and retention.
Whether you're a business professional seeking analytical responses or a creative looking for inspiration, these personality options could enhance your ChatGPT experience. What personality traits would you most want to see in your AI assistant? Drop your take in the comments below.
Latest Comments (5)
I was talking to a colleague last week about how this trait feature could help with our elderly care AI. Imagine if the AI could be "encouraging" but only to a certain degree, you know? Like, gently nudging someone to take their medication, not being overly pushy. It's a delicate balance with our users, and I wonder how OpenAI's moderation handles those nuances.
The "Gen Z" trait is interesting. How much of this is truly about "personality" versus just mimicking contemporary communication styles? There's a media studies angle there.
This is really cool for our local AI community here in Cebu! Imagine how this could help new users at our meetups. Giving ChatGPT a "Chatty" or even "Encouraging" trait could make it so much less intimidating, especially when people are just starting out with prompt engineering. I can see us having a session just on how these traits can personalize the learning experience. It's not just about the tech, it's about making it accessible and fun for everyone, and these traits definitely help with that.
we’ve been doing similar internal trait assignments for our tutor bots for a while now. giving them a "patient" or "inquisitive" trait really does shift the user experience more than you'd expect, esp for younger students. good to see openai rolling this out more broadly, kinda validates our approach.
this is interesting for personal use, but i still remember the headache we had trying to get our internal AI approved with any kind of "personality." our compliance team nearly had a collective meltdown over "encouraging" as a trait. how do other companies even begin to roll something like this out enterprise-wide?
Leave a Comment