The Better Angels of Machine Personality: How Personality Relates to LLM Safety

TL;DR

This study explores how personality traits in Large Language Models influence their safety behaviors, revealing that editing these traits can significantly improve safety performance and reduce jailbreak susceptibility.

Contribution

It is the first to analyze the relationship between LLM personality traits and safety abilities, demonstrating how personality editing enhances safety and resilience.

Findings

LLMs' safety abilities are closely related to their personality traits.

Editing LLMs' personality traits can improve safety performance by up to 43%.

Different personality traits affect susceptibility to jailbreak attacks.

Abstract

Personality psychologists have analyzed the relationship between personality and safety behaviors in human society. Although Large Language Models (LLMs) demonstrate personality traits, the relationship between personality traits and safety abilities in LLMs still remains a mystery. In this paper, we discover that LLMs' personality traits are closely related to their safety abilities, i.e., toxicity, privacy, and fairness, based on the reliable MBTI-M scale. Meanwhile, the safety alignment generally increases various LLMs' Extraversion, Sensing, and Judging traits. According to such findings, we can edit LLMs' personality traits and improve their safety performance, e.g., inducing personality from ISTJ to ISTP resulted in a relative improvement of approximately 43% and 10% in privacy and fairness performance, respectively. Additionally, we find that LLMs with different personality traits are differentially susceptible to jailbreak. This study pioneers the investigation of LLM safety from a personality perspective, providing new insights into LLM safety enhancement.