@inproceedings{kabir-etal-2026-press,
title = "{PR}e{SS}: An Automated Black-Box Framework for Evaluating Political Stance Stability in {LLM}s",
author = "Kabir, Shariar and
Dong, Yue and
Esterling, Kevin",
editor = "Afli, Haithem and
Bouamor, Houda and
Zaghouani, Wajdi and
Ghannay, Sahar and
Hossain, Shehenaz",
booktitle = "Proceedings of the 3rd Workshop on Natural Language Processing for Political Sciences ({P}olitical{NLP} 2026)",
month = may,
year = "2026",
address = "Palma, Mallorca (Spain)",
publisher = "ELRA Language Resources Association (ELRA)",
url = "https://aclanthology.org/2026.politicalnlp-1.26/",
doi = "10.63317/35d8pipu4ofv",
pages = "234--247",
abstract = "Existing evaluations of political bias in large language models (LLMs) typically classify outputs as left- or right-leaning. We extend this perspective by examining how ideological tendencies vary across topics and how consistently models maintain their positions, a property we refer to as stability. To capture this dimension, we propose PReSS (Political Response Stability under Stress), an automated black-box framework that evaluates LLMs by jointly considering model and topic context, categorizing responses into four stance types: stable-left, unstable-left, stable-right, and unstable-right. Applying PReSS to 9 widely used LLMs across 19 political topics reveals substantial variation in stance stability; for instance, a model that is left-leaning overall can exhibit stable-right behavior on certain topics. This highlights the importance of topic-aware and fine-grained evaluation of political ideologies of LLMs. Moreover, stability has practical implications for controlled generation and model alignment: interventions such as debiasing or ideology reversal should explicitly account for stance stability. Our empirical analyses reveal that when models are prompted or fine-tuned to adopt the opposite ideology, unstable topic stances are more likely to change, whereas stable ones resist modification. Thus, treating stability as a moderating factor provides a principled foundation for understanding, evaluating, and guiding interventions in politically sensitive model behavior."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="kabir-etal-2026-press">
<titleInfo>
<title>PReSS: An Automated Black-Box Framework for Evaluating Political Stance Stability in LLMs</title>
</titleInfo>
<name type="personal">
<namePart type="given">Shariar</namePart>
<namePart type="family">Kabir</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yue</namePart>
<namePart type="family">Dong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kevin</namePart>
<namePart type="family">Esterling</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 3rd Workshop on Natural Language Processing for Political Sciences (PoliticalNLP 2026)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Haithem</namePart>
<namePart type="family">Afli</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Houda</namePart>
<namePart type="family">Bouamor</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wajdi</namePart>
<namePart type="family">Zaghouani</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sahar</namePart>
<namePart type="family">Ghannay</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shehenaz</namePart>
<namePart type="family">Hossain</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>ELRA Language Resources Association (ELRA)</publisher>
<place>
<placeTerm type="text">Palma, Mallorca (Spain)</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Existing evaluations of political bias in large language models (LLMs) typically classify outputs as left- or right-leaning. We extend this perspective by examining how ideological tendencies vary across topics and how consistently models maintain their positions, a property we refer to as stability. To capture this dimension, we propose PReSS (Political Response Stability under Stress), an automated black-box framework that evaluates LLMs by jointly considering model and topic context, categorizing responses into four stance types: stable-left, unstable-left, stable-right, and unstable-right. Applying PReSS to 9 widely used LLMs across 19 political topics reveals substantial variation in stance stability; for instance, a model that is left-leaning overall can exhibit stable-right behavior on certain topics. This highlights the importance of topic-aware and fine-grained evaluation of political ideologies of LLMs. Moreover, stability has practical implications for controlled generation and model alignment: interventions such as debiasing or ideology reversal should explicitly account for stance stability. Our empirical analyses reveal that when models are prompted or fine-tuned to adopt the opposite ideology, unstable topic stances are more likely to change, whereas stable ones resist modification. Thus, treating stability as a moderating factor provides a principled foundation for understanding, evaluating, and guiding interventions in politically sensitive model behavior.</abstract>
<identifier type="citekey">kabir-etal-2026-press</identifier>
<identifier type="doi">10.63317/35d8pipu4ofv</identifier>
<location>
<url>https://aclanthology.org/2026.politicalnlp-1.26/</url>
</location>
<part>
<date>2026-05</date>
<extent unit="page">
<start>234</start>
<end>247</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T PReSS: An Automated Black-Box Framework for Evaluating Political Stance Stability in LLMs
%A Kabir, Shariar
%A Dong, Yue
%A Esterling, Kevin
%Y Afli, Haithem
%Y Bouamor, Houda
%Y Zaghouani, Wajdi
%Y Ghannay, Sahar
%Y Hossain, Shehenaz
%S Proceedings of the 3rd Workshop on Natural Language Processing for Political Sciences (PoliticalNLP 2026)
%D 2026
%8 May
%I ELRA Language Resources Association (ELRA)
%C Palma, Mallorca (Spain)
%F kabir-etal-2026-press
%X Existing evaluations of political bias in large language models (LLMs) typically classify outputs as left- or right-leaning. We extend this perspective by examining how ideological tendencies vary across topics and how consistently models maintain their positions, a property we refer to as stability. To capture this dimension, we propose PReSS (Political Response Stability under Stress), an automated black-box framework that evaluates LLMs by jointly considering model and topic context, categorizing responses into four stance types: stable-left, unstable-left, stable-right, and unstable-right. Applying PReSS to 9 widely used LLMs across 19 political topics reveals substantial variation in stance stability; for instance, a model that is left-leaning overall can exhibit stable-right behavior on certain topics. This highlights the importance of topic-aware and fine-grained evaluation of political ideologies of LLMs. Moreover, stability has practical implications for controlled generation and model alignment: interventions such as debiasing or ideology reversal should explicitly account for stance stability. Our empirical analyses reveal that when models are prompted or fine-tuned to adopt the opposite ideology, unstable topic stances are more likely to change, whereas stable ones resist modification. Thus, treating stability as a moderating factor provides a principled foundation for understanding, evaluating, and guiding interventions in politically sensitive model behavior.
%R 10.63317/35d8pipu4ofv
%U https://aclanthology.org/2026.politicalnlp-1.26/
%U https://doi.org/10.63317/35d8pipu4ofv
%P 234-247
Markdown (Informal)
[PReSS: An Automated Black-Box Framework for Evaluating Political Stance Stability in LLMs](https://aclanthology.org/2026.politicalnlp-1.26/) (Kabir et al., PoliticalNLP 2026)
ACL