@inproceedings{dong-etal-2026-persistent,
title = "{I} Am Not Them: Persistent Outgroup Bias in Large Language Models Arising from Social Identity Persona Setting",
author = "Dong, Wenchao and
Zhunis, Assem and
Jeong, Dongyoung and
Chin, Hyojin and
Han, Jiyoung and
Cha, Meeyoung",
editor = "Piperidis, Stelios and
Bel, N{\'u}ria and
van den Heuvel, Henk and
Ide, Nancy and
Krek, Simon and
Toral, Antonio",
booktitle = "Proceedings of the Fifteenth Language Resources and Evaluation Conference",
month = may,
year = "2026",
address = "Palma de Mallorca, Spain",
publisher = "ELRA Language Resource Association",
url = "https://aclanthology.org/2026.lrec-1.217/",
doi = "10.63317/2hn5gs6yh5m2",
pages = "2772--2786",
abstract = "This research examines how large language models internalize social identities assigned through targeted prompts. Guided by social identity theory, we investigate whether and how these identity assignments cause AI systems to differentiate between ``we'' (the ingroup) and ``they'' (the outgroup). We demonstrate that self-categorization of social identity leads to both ingroup favoritism and outgroup bias, with the latter manifesting as strongly as the former. This finding is significant given the fundamental role of outgroup bias in driving intergroup prejudice and discrimination as documented in social psychology. We further propose a strategic intervention to mitigate such bias by guiding language models to adopt the identity of the initially disfavored group. This method, validated across both political and gender domains, exposes a critical dual function of group alignment: adopting one social identity inherently alters the model{'}s stance toward outgroups, effectively neutralizing pre-existing biases. Our work shows that understanding human-like AI behaviors is a critical prerequisite to building more balanced and socially responsible technology."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="dong-etal-2026-persistent">
<titleInfo>
<title>I Am Not Them: Persistent Outgroup Bias in Large Language Models Arising from Social Identity Persona Setting</title>
</titleInfo>
<name type="personal">
<namePart type="given">Wenchao</namePart>
<namePart type="family">Dong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Assem</namePart>
<namePart type="family">Zhunis</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dongyoung</namePart>
<namePart type="family">Jeong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hyojin</namePart>
<namePart type="family">Chin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiyoung</namePart>
<namePart type="family">Han</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Meeyoung</namePart>
<namePart type="family">Cha</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-05</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the Fifteenth Language Resources and Evaluation Conference</title>
</titleInfo>
<name type="personal">
<namePart type="given">Stelios</namePart>
<namePart type="family">Piperidis</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Núria</namePart>
<namePart type="family">Bel</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Henk</namePart>
<namePart type="family">van den Heuvel</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nancy</namePart>
<namePart type="family">Ide</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Simon</namePart>
<namePart type="family">Krek</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Antonio</namePart>
<namePart type="family">Toral</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>ELRA Language Resource Association</publisher>
<place>
<placeTerm type="text">Palma de Mallorca, Spain</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>This research examines how large language models internalize social identities assigned through targeted prompts. Guided by social identity theory, we investigate whether and how these identity assignments cause AI systems to differentiate between “we” (the ingroup) and “they” (the outgroup). We demonstrate that self-categorization of social identity leads to both ingroup favoritism and outgroup bias, with the latter manifesting as strongly as the former. This finding is significant given the fundamental role of outgroup bias in driving intergroup prejudice and discrimination as documented in social psychology. We further propose a strategic intervention to mitigate such bias by guiding language models to adopt the identity of the initially disfavored group. This method, validated across both political and gender domains, exposes a critical dual function of group alignment: adopting one social identity inherently alters the model’s stance toward outgroups, effectively neutralizing pre-existing biases. Our work shows that understanding human-like AI behaviors is a critical prerequisite to building more balanced and socially responsible technology.</abstract>
<identifier type="citekey">dong-etal-2026-persistent</identifier>
<identifier type="doi">10.63317/2hn5gs6yh5m2</identifier>
<location>
<url>https://aclanthology.org/2026.lrec-1.217/</url>
</location>
<part>
<date>2026-05</date>
<extent unit="page">
<start>2772</start>
<end>2786</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T I Am Not Them: Persistent Outgroup Bias in Large Language Models Arising from Social Identity Persona Setting
%A Dong, Wenchao
%A Zhunis, Assem
%A Jeong, Dongyoung
%A Chin, Hyojin
%A Han, Jiyoung
%A Cha, Meeyoung
%Y Piperidis, Stelios
%Y Bel, Núria
%Y van den Heuvel, Henk
%Y Ide, Nancy
%Y Krek, Simon
%Y Toral, Antonio
%S Proceedings of the Fifteenth Language Resources and Evaluation Conference
%D 2026
%8 May
%I ELRA Language Resource Association
%C Palma de Mallorca, Spain
%F dong-etal-2026-persistent
%X This research examines how large language models internalize social identities assigned through targeted prompts. Guided by social identity theory, we investigate whether and how these identity assignments cause AI systems to differentiate between “we” (the ingroup) and “they” (the outgroup). We demonstrate that self-categorization of social identity leads to both ingroup favoritism and outgroup bias, with the latter manifesting as strongly as the former. This finding is significant given the fundamental role of outgroup bias in driving intergroup prejudice and discrimination as documented in social psychology. We further propose a strategic intervention to mitigate such bias by guiding language models to adopt the identity of the initially disfavored group. This method, validated across both political and gender domains, exposes a critical dual function of group alignment: adopting one social identity inherently alters the model’s stance toward outgroups, effectively neutralizing pre-existing biases. Our work shows that understanding human-like AI behaviors is a critical prerequisite to building more balanced and socially responsible technology.
%R 10.63317/2hn5gs6yh5m2
%U https://aclanthology.org/2026.lrec-1.217/
%U https://doi.org/10.63317/2hn5gs6yh5m2
%P 2772-2786
Markdown (Informal)
[I Am Not Them: Persistent Outgroup Bias in Large Language Models Arising from Social Identity Persona Setting](https://aclanthology.org/2026.lrec-1.217/) (Dong et al., LREC 2026)
ACL