The Other Side of the Coin: Exploring Fairness in Retrieval-Augmented Generation

2025-04-11

AI Summary

This work investigates the impact of retrieval-augmented generation (RAG) on the fairness of large language models (LLMs), revealing for the first time that integrating RAG into small-scale models (<8B parameters) significantly exacerbates fairness disparities. To address this, we propose a dual-path mitigation framework: FairFT—a retriever-LLM fair alignment fine-tuning method leveraging contrastive learning and preference optimization—and FairFilter—a post-retrieval bias filtering mechanism combining rule-based and model-based techniques. Evaluated on multiple real-world fairness benchmarks using open-source models including LLaMA3 and Mistral, our approach achieves an average 23.6% improvement in fairness metrics while preserving 98.4% of original task performance. The core contribution lies in identifying the “fairness paradox” of RAG in small models and establishing the first end-to-end fairness optimization paradigm specifically designed for the RAG paradigm.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Ethics — Bias, Fairness, Transparency & Privacy
• Computer Vision: Bias, Fairness & Privacy

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Social Networks and Social Media: Fairness and bias in social network and social media analysis

Abstract

Retrieval-Augmented Generation (RAG) enhances Large Language Models (LLMs) by retrieving relevant document from external knowledge sources. By referencing this external knowledge, RAG effectively reduces the generation of factually incorrect content and addresses hallucination issues within LLMs. Recently, there has been growing attention to improving the performance and efficiency of RAG systems from various perspectives. While these advancements have yielded significant results, the application of RAG in domains with considerable societal implications raises a critical question about fairness: What impact does the introduction of the RAG paradigm have on the fairness of LLMs? To address this question, we conduct extensive experiments by varying the LLMs, retrievers, and retrieval sources. Our experimental analysis reveals that the scale of the LLMs plays a significant role in influencing fairness outcomes within the RAG framework. When the model scale is smaller than 8B, the integration of retrieval mechanisms often exacerbates unfairness in small-scale LLMs (e.g., LLaMA3.2-1B, Mistral-7B, and LLaMA3-8B). To mitigate the fairness issues introduced by RAG for small-scale LLMs, we propose two approaches, FairFT and FairFilter. Specifically, in FairFT, we align the retriever with the LLM in terms of fairness, enabling it to retrieve documents that facilitate fairer model outputs. In FairFilter, we propose a fairness filtering mechanism to filter out biased content after retrieval. Finally, we validate our proposed approaches on real-world datasets, demonstrating their effectiveness in improving fairness while maintaining performance.

Problem

• Investigates fairness impact of RAG on LLMs
• Examines unfairness in small-scale LLMs with RAG
• Proposes FairFT and FairFilter to mitigate RAG fairness issues

Innovation

• Aligns retriever with LLM for fairness
• Filters biased content post-retrieval
• Validates on real-world datasets effectively