Pith. sign in

REVIEW 5 cited by

Bias Amplification in RAG: Poisoning Knowledge Retrieval to Steer LLMs

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2506.11415 v1 pith:AXP3J7RD submitted 2025-06-13 cs.LG cs.CLcs.CR

Bias Amplification in RAG: Poisoning Knowledge Retrieval to Steer LLMs

classification cs.LG cs.CLcs.CR
keywords modelattackbiasbiaseslanguagesystemamplifyattacks
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

In Large Language Models, Retrieval-Augmented Generation (RAG) systems can significantly enhance the performance of large language models by integrating external knowledge. However, RAG also introduces new security risks. Existing research focuses mainly on how poisoning attacks in RAG systems affect model output quality, overlooking their potential to amplify model biases. For example, when querying about domestic violence victims, a compromised RAG system might preferentially retrieve documents depicting women as victims, causing the model to generate outputs that perpetuate gender stereotypes even when the original query is gender neutral. To show the impact of the bias, this paper proposes a Bias Retrieval and Reward Attack (BRRA) framework, which systematically investigates attack pathways that amplify language model biases through a RAG system manipulation. We design an adversarial document generation method based on multi-objective reward functions, employ subspace projection techniques to manipulate retrieval results, and construct a cyclic feedback mechanism for continuous bias amplification. Experiments on multiple mainstream large language models demonstrate that BRRA attacks can significantly enhance model biases in dimensions. In addition, we explore a dual stage defense mechanism to effectively mitigate the impacts of the attack. This study reveals that poisoning attacks in RAG systems directly amplify model output biases and clarifies the relationship between RAG system security and model fairness. This novel potential attack indicates that we need to keep an eye on the fairness issues of the RAG system.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 5 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. SoK: Colluding Adversaries in Machine Learning Pipelines

    cs.CR 2026-06 unverdicted novelty 7.0

    The paper introduces a framework for collusion between train- and inference-time adversaries in ML pipelines, proposes a guideline for conjecturing collusion potential, explains prior work, and empirically validates f...

  2. Epistemic Bias Injection: Manipulating LLM Opinion via Selective Context Retrieval

    cs.CR 2025-11 conditional novelty 6.0

    Factually correct but one-sided passages injected into a RAG database can shift LLM answers toward an attacker's stance, and a PCA/KL-based filter (BiasDef) reduces that shift.

  3. Large Language Models in Misinformation Ecosystems: Misuse, Defense, and Vulnerability

    cs.CR 2026-07 conditional novelty 5.0

    A role-layer survey unifies LLM misuse, LLM-based defense, and LLM-centric verification vulnerabilities across content, social, evidence, and workflow layers, then lists three open challenges.

  4. Retrieval-Augmented Generation Must Move Beyond Factual Grounding to Represent Diverse Opinions

    cs.AI 2026-04 unverdicted novelty 5.0

    Opinion-aware RAG with LLM opinion extraction and entity-linked graphs improves retrieval diversity by 26-42% over factual baselines on e-commerce forum data.

  5. Retrieval-Augmented Generation Must Move Beyond Factual Grounding to Represent Diverse Opinions

    cs.AI 2026-04 unverdicted novelty 5.0

    RAG systems optimize for factual certainty and ignore opinion diversity; O-RAG reduces Wasserstein distance to corpus sentiment and is preferred by humans 79% of the time.