Challenge: Language model alignment research often attempts to ensure that models are helpful and harmless, but can obscure how improving one aspect might impact the other.
Approach: They analyze the relationship between truthfulness and political bias in language models.
Outcome: The results show that optimizing models for truthfulness results in a left-leaning political bias.

Similar Papers

Whose Emotions and Moral Sentiments do Language Models Reflect? (2024.findings-acl)

Copied to clipboard

Challenge: Existing research has focused on positional alignment, which measures how closely the models mimic the opinions and stances of different social groups.
Approach: They define the problem of affective alignment, which measures how LMs’ emotional and moral tone represents those of different groups.
Outcome: The results show that the models represent the perspectives of some social groups better than others, suggesting a systemic bias within LMs.
PolBiX: Detecting LLMs’ Political Bias in Fact-Checking through X-phemisms (2025.findings-emnlp)

Copied to clipboard

Challenge: a few models show tendencies of political bias, but this is not mitigated by explicitly calling for objectivism in prompts.
Approach: They investigate political bias by exchanging words with euphemisms or dysphemismas in German claims.
Outcome: The proposed model shows that political bias influences truthfulness assessment more than political leaning .
The Hidden Bias: A Study on Explicit and Implicit Political Stereotypes in Large Language Models (2026.findings-eacl)

Copied to clipboard

Challenge: Large Language Models (LLMs) are increasingly integral to information dissemination and decision-making processes.
Approach: They investigate political bias and stereotype propagation across eight prominent LLMs using the two-dimensional Political Compass Test.
Outcome: The political bias and stereotype propagation of large language models is investigated using the two-dimensional Political Compass Test (PCT) key findings reveal a left-leaning political alignment across all investigated models.
From Pretraining Data to Language Models to Downstream Tasks: Tracking the Trails of Political Biases Leading to Unfair NLP Models (2023.acl-long)

Copied to clipboard

Challenge: Hundreds of studies have highlighted ethical issues in NLP models .
Approach: They propose to measure media biases in LMs trained on diverse data sources . they focus on hate speech and misinformation detection .
Outcome: The proposed methods quantify the fairness of downstream NLP models trained on politically biased LMs.
Analyzing Political Bias in LLMs via Target-Oriented Sentiment Classification (2025.findings-acl)

Copied to clipboard

Challenge: Existing methods to analyze political biases rely on small-size intermediate tasks and the LLMs themselves.
Approach: They propose an entropy-based inconsistency metric to encode political biases . they insert 1319 demographically and politically diverse politician names in 450 political sentences .
Outcome: The proposed method combines high accuracy with a correct understanding of the candidate candidate.
Democratic or Authoritarian? Probing a New Dimension of Political Biases in Large Language Models (2026.eacl-long)

Copied to clipboard

Challenge: Prior work on LLM biases focused on socio-demographic and left–right political dimensions, but little attention has been paid to how they align with broader geopolitical value systems.
Approach: They propose a method to assess how LLMs align with broader geopolitical value systems, particularly the democracy–authoritarianism spectrum.
Outcome: The proposed method combines the F-scale, FavScore and role-model probing to assess which figures are cited as general role models by LLMs.
How Gender Interacts with Political Values: A Case Study on Czech BERT Models (2024.lrec-main)

Copied to clipboard

Challenge: Neural language models are trained on large text corpora that contain value-burdened content and often capture undesirable biases, which the models reflect.
Approach: They propose a method to measure the model's perceived political values by comparing Czech with a representative value survey.
Outcome: The proposed method does not assign statement probability following value-driven reasoning and there is no systematic difference between feminine and masculine sentences.
Measuring Political Bias in Large Language Models: What Is Said and How It Is Said (2024.acl-long)

Copied to clipboard

Challenge: Existing benchmarks and measures focus on gender and racial biases, but political bias exists in LLMs and can lead to polarization and other harms in downstream applications.
Approach: They propose to analyze the content and style of LLMs generated by political issues and propose a framework that can be scalable to other topics.
Outcome: The proposed framework is easily scalable to other topics and is explainable.
Bias in the East, Bias in the West: A Bilingual Analysis of LLM Political Bias on U.S.- and China-Related Issues (2026.findings-eacl)

Copied to clipboard

Challenge: Large language models (LLMs) can exhibit political biases, which creates a risk of undue influence on LLM users and public opinion.
Approach: They use a dataset of 36k real-time test prompts to measure LLM political bias on U.S. and Chinese issues.
Outcome: The proposed model origin and prompt language influence bias on 60 political issues.
Only a Little to the Left: A Theory-grounded Measure of Political Bias in Large Language Models (2025.acl-long)

Copied to clipboard

Challenge: Political biases in language models can affect performance in many applications . political biased models are often left-leaning, but are generally more left- leaning for instruction-tuned models .
Approach: They propose to use the Political Compass Test to measure political bias in language models . they use survey-based evaluation tools to test prompts and classify their political stances .
Outcome: The proposed model is based on the Political Compass Test, but is not scientifically valid.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations