Assessing GPT's Bias Towards Stigmatized Social Groups: An Intersectional Case Study on Nationality Prejudice and Psychophobia

Afifah Kashif, Heer Patel
5/15/2025
cs.CL

No confident code implementation found for this paper.

Abstract

Recent studies have separately highlighted significant biases within foundational large language models (LLMs) against certain nationalities and stigmatized social groups. This research investigates the ethical implications of these biases intersecting with outputs of widely-used GPT-3.5/4/4o LLMS. Through structured prompt series, we evaluate model responses to several scenarios involving American and North Korean nationalities with various mental disabilities. Findings reveal significant discrepancies in empathy levels with North Koreans facing greater negative bias, particularly when mental disability is also a factor. This underscores the need for improvements in LLMs designed with a nuanced understanding of intersectional identity.

View on arXivView PDF

Code Implementations

No confident code match yet

We couldn't find an author-owned or strongly-evidenced community implementation for this paper. Any repos shown below are weak matches — verify before relying on them.

No code implementations found yet.

Know of an implementation? Let us know in the comments below!

Similar Papers

Identifying Features Associated with Bias Against 93 Stigmatized Groups in Language Models and Guardrail Model Safety Mitigation

86% match1 repo12025

Large Language Models Portray Socially Subordinate Groups as More Homogeneous, Consistent with a Bias Observed in Humans

86% match2024

Bias Against 93 Stigmatized Groups in Masked Language Models and Downstream Sentiment Classification Tasks

85% match2023

Investigating Intersectional Bias in Large Language Models using Confidence Disparities in Coreference Resolution

84% match2025

Evaluating Biased Attitude Associations of Language Models in an Intersectional Context

84% match1 repo32023

Laissez-Faire Harms: Algorithmic Biases in Generative Language Models

84% match2024

Cite this paper

@article{kashif2025assessing,
  title  = {Assessing GPT's Bias Towards Stigmatized Social Groups: An Intersectional Case Study on Nationality Prejudice and Psychophobia},
  author = {Afifah Kashif and Heer Patel},
  year   = {2025},
  eprint = {2505.17045},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2505.17045v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2505.17045.svg)](https://codeofpaper.com/paper/2505.17045)

Discussion