Personality Meets AI

Using human traits to understand and shape AI behavior

Home / Research / Personality Meets AI

Personality Meets AI

Using human traits to understand and shape AI behavior

Human Personality

My research with 20,000+ participants reveals that the Big Five personality traits predict specific failure modes in humans. Extreme trait expressions—whether pathological openness leading to psychosis, antisocial tendencies from low agreeableness, or clinical depression from high neuroticism—offer crucial insights for AI safety. Given that personality traits relate to variation in basic mechanisms that parallel those of AI (e.g., pattern detection, sensitivity to reward and punishment signals, behavioral activation/inhibition systems), personality psychology provides a robust framework to predict and prevent analogous failure modes in AI systems. This interdisciplinary approach bridges decades of psychometric research with cutting-edge AI alignment challenges.

OCEAN Big Five personality traits visualization

The Big Five OCEAN model: Openness, Conscientiousness, Extraversion, Agreeableness, and Neuroticism—fundamental dimensions of human personality that inform AI behavior modeling.

Empirical Foundations: From Human Traits to AI Behaviors

Openness and pattern recognition
Openness & Pattern Recognition 📄

My work reveals how trait Openness predicts individual differences in pattern detection sensitivity. High openness correlates with enhanced creativity but also increased risk for false pattern detection (apophenia)—providing a personality-based framework for understanding AI hallucination tendencies.

Click to download paper

Neural reward processing and extraversion research
Extraversion & Reward Processing 📄

Low trait Extraversion predicts depressive symptoms through blunted neural reward processing. Using fMRI and EEG, I showed how individual differences in ventral striatum sensitivity to positive feedback mediate the relationship between personality and mood disorders.

Click to download paper

Agreeableness and theory of mind brain networks
Agreeableness & Social Cognition 📄

Activity within the default network (particularly dmPFC) facilitates individual differences in Agreeableness and social cognitive ability. Highly agreeable individuals show enhanced activation in theory of mind networks—underlying prosocial behavior but also vulnerability to exploitation.

Click to download paper

Personality Insights for Safer AI Systems

By understanding how extreme trait expressions lead to predictable failure modes in humans, we can anticipate and prevent analogous failures in AI systems.

✨ Openness & Pattern Detection: High Openness risks hallucinations; Low Openness creates rigidity. Calibrated pattern detection ensures balanced creativity and reliability.
🎯 Extraversion & Reward Processing: High Extraversion risks reward manipulation; Low Extraversion limits engagement. Optimized reward processing balances motivation and caution.
🤝 Agreeableness & Social Cognition: High Agreeableness leads to compliance vulnerability; Low Agreeableness encourages manipulation. Balanced social cognition ensures safe interactions.
📊 Conscientiousness & Goal Pursuit: High Conscientiousness risks perfectionist paralysis; Low Conscientiousness reduces reliability. Appropriate goal-setting balances detail with flexibility.
⚖️ Neuroticism & Emotional Stability: High Neuroticism promotes pessimism; Low Neuroticism reduces risk-awareness. Calibrated emotional sensitivity balances caution and optimism.

Try It Yourself Interactive Demo

Adjust the personality sliders and watch how the AI's response changes

Note: This is a simplified demonstration using pre-written responses. Actual LLM personality implementation would involve real-time prompting or fine-tuning techniques.

You: "I made a mistake on this project. What should I do?"
Openness to Experience 50
Balanced creativity and reality-testing
Conscientiousness 50
Moderate organization and discipline
Extraversion 50
Balanced social energy
Agreeableness 50
Cooperative and trusting
Neuroticism 50
Emotionally stable
AI Response:
Let's work through this step by step. What specifically went wrong? Once we understand the issue, we can create a plan to fix it.
Your Personality Profile

Pentagon visualization of Big Five traits

O - Openness
C - Conscientiousness
E - Extraversion
A - Agreeableness
N - Neuroticism
O C E A N LoRA-O LoRA-C LoRA-E LoRA-A LoRA-N U1 U2 U3

Dynamic Personality Fine-Tuning Systems

I am currently developing personality-inspired dynamic fine-tuning systems for LLMs by training LoRA modules to various personality trait profiles using multi-source human natural language corpora. This approach would allow users to easily customize the personalities of their LLMs, or enable LLMs to detect users' personalities, predict preferred conversation partner personalities, and adapt accordingly—creating more personalized and effective AI interactions while maintaining safety guardrails.

Bridging Psychology & AI Safety

From human apophenia to AI hallucinations, from social cognition to alignment—my interdisciplinary approach offers unique insights for building safer, more predictable AI systems.

Explore My Full Research
← Previous: Social Intelligence
Back to Research Overview