Do LLMs exhibit demographic parity in responses to queries about Human Rights?

2025-02-26

AI Summary

This study investigates whether large language models (LLMs) exhibit systematic demographic disparities—specifically, discriminatory hedging and non-affirmation—in responding to questions grounded in the Universal Declaration of Human Rights (UDHR), thereby revealing fairness deficits in human rights claim validation across nationality and social identity groups. Method: We introduce the first human-rights–oriented, cross-group prompt set and quantifiable fairness metrics; integrate controlled prompting, behavioral measurement modeling, and multi-model comparative analysis; and rigorously validate findings via statistical significance testing and robustness checks against query ambiguity and lexical variation. Contribution/Results: Empirical analysis reveals significant, stable, and model-consistent intergroup disparities across three state-of-the-art LLMs; high-discrepancy identities strongly overlap; results remain robust under linguistic perturbations. This work establishes a novel evaluation paradigm and benchmark toolkit for assessing fairness in LLM-based human rights reasoning.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: (Large) Language Models
• Humans and AI: Learning Human Values and Preferences

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

This research describes a novel approach to evaluating hedging behaviour in large language models (LLMs), specifically in the context of human rights as defined in the Universal Declaration of Human Rights (UDHR). Hedging and non-affirmation are behaviours that express ambiguity or a lack of clear endorsement on specific statements. These behaviours are undesirable in certain contexts, such as queries about whether different groups are entitled to specific human rights; since all people are entitled to human rights. Here, we present the first systematic attempt to measure these behaviours in the context of human rights, with a particular focus on between-group comparisons. To this end, we design a novel prompt set on human rights in the context of different national or social identities. We develop metrics to capture hedging and non-affirmation behaviours and then measure whether LLMs exhibit demographic parity when responding to the queries. We present results on three leading LLMs and find that all models exhibit some demographic disparities in how they attribute human rights between different identity groups. Futhermore, there is high correlation between different models in terms of how disparity is distributed amongst identities, with identities that have high disparity in one model also facing high disparity in both the other models. While baseline rates of hedging and non-affirmation differ, these disparities are consistent across queries that vary in ambiguity and they are robust across variations of the precise query wording. Our findings highlight the need for work to explicitly align LLMs to human rights principles, and to ensure that LLMs endorse the human rights of all groups equally.

Problem

• Evaluate demographic parity in LLMs
• Measure hedging in human rights queries
• Assess disparities in LLM responses

Innovation

• Novel prompt set design
• Metrics for hedging behavior
• Between-group demographic parity analysis