Sometimes the Model doth Preach: Quantifying Religious Bias in Open LLMs through Demographic Analysis in Asian Nations

2025-03-10

AI Summary

This study investigates latent religious biases in open-source large language models (LLMs) within Asian sociocultural contexts—particularly India. To address this, we propose a Hamming-distance-based metric for quantifying alignment between model outputs and human responses from local populations, enabling the first systematic, cross-cultural assessment of religious tolerance and identity-related biases in models such as Llama and Mistral. Our analysis reveals that mainstream open-source LLMs consistently diverge from empirically observed pluralistic religious attitudes across multiple Asian countries, instead converging toward the perspectives of dominant demographic groups—indicating cultural homogenization and inadequate regional grounding. Beyond exposing risks of epistemic and cultural hegemony in non-Western settings, this work introduces a reproducible, scalable diagnostic framework for religion-sensitive AI bias evaluation. It advances global AI fairness assessment and governance by establishing a novel, context-aware paradigm for auditing sociocultural alignment in foundation models.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Philosophy and Ethics of AI: Bias, Fairness & Equity
• Natural Language Processing: Ethics — Bias, Fairness, Transparency & Privacy

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Search and Retrieval-Augmented AI: Large language models for search
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Large Language Models (LLMs) are capable of generating opinions and propagating bias unknowingly, originating from unrepresentative and non-diverse data collection. Prior research has analysed these opinions with respect to the West, particularly the United States. However, insights thus produced may not be generalized in non-Western populations. With the widespread usage of LLM systems by users across several different walks of life, the cultural sensitivity of each generated output is of crucial interest. Our work proposes a novel method that quantitatively analyzes the opinions generated by LLMs, improving on previous work with regards to extracting the social demographics of the models. Our method measures the distance from an LLM's response to survey respondents, through Hamming Distance, to infer the demographic characteristics reflected in the model's outputs. We evaluate modern, open LLMs such as Llama and Mistral on surveys conducted in various global south countries, with a focus on India and other Asian nations, specifically assessing the model's performance on surveys related to religious tolerance and identity. Our analysis reveals that most open LLMs match a single homogeneous profile, varying across different countries/territories, which in turn raises questions about the risks of LLMs promoting a hegemonic worldview, and undermining perspectives of different minorities. Our framework may also be useful for future research investigating the complex intersection between training data, model architecture, and the resulting biases reflected in LLM outputs, particularly concerning sensitive topics like religious tolerance and identity.

Problem

• Quantify religious bias in open LLMs using demographic analysis in Asia.
• Assess LLM cultural sensitivity and bias in non-Western contexts.
• Develop a method to measure LLM outputs' alignment with diverse demographics.

Innovation

• Quantitative analysis of LLM opinions using Hamming Distance
• Focus on religious bias in Asian nations via demographic analysis
• Evaluation of open LLMs like Llama and Mistral on global surveys