Gender Bias in LLM-generated Interview Responses

TL;DR

This paper systematically examines gender bias in interview responses generated by LLMs, revealing persistent biases aligned with stereotypes and job dominance, emphasizing the need for bias mitigation.

Contribution

It provides a comprehensive evaluation of gender bias across multiple LLMs, question types, and jobs, offering insights into bias patterns and implications for fair AI use.

Findings

Gender bias is consistent across models and questions.

Bias aligns with gender stereotypes and job dominance.

Highlights the importance of mitigating bias in LLM applications.

Abstract

LLMs have emerged as a promising tool for assisting individuals in diverse text-generation tasks, including job-related texts. However, LLM-generated answers have been increasingly found to exhibit gender bias. This study evaluates three LLMs (GPT-3.5, GPT-4, Claude) to conduct a multifaceted audit of LLM-generated interview responses across models, question types, and jobs, and their alignment with two gender stereotypes. Our findings reveal that gender bias is consistent, and closely aligned with gender stereotypes and the dominance of jobs. Overall, this study contributes to the systematic examination of gender bias in LLM-generated interview responses, highlighting the need for a mindful approach to mitigate such biases in related applications.