# Epistemic Bias Injection: Manipulating LLM Opinion via Selective Context Retrieval

arXiv: https://arxiv.org/abs/2512.00804 · HTML: https://mlsanity.com/paper/2512.00804/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Placed in Privacy and security of RAG, one of 15 directions in Retrieval-augmented generation.
- Builds on 34 of its 52 references that are in the map.
- First posted on arXiv 30 Nov 2025.

## Placed in

- Retrieval-augmented generation / [Privacy and security of RAG](https://mlsanity.com/rag/privacy-and-security-of-rag/)

## Builds on (most cited first, 30 of 34)

- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [Evaluating Large Language Models Trained on Code](https://mlsanity.com/paper/2107.03374/) (2021)
- [Efficient Memory Management for Large Language Model Serving with PagedAttention](https://mlsanity.com/paper/2309.06180/) (2023)
- [Dense Passage Retrieval for Open-Domain Question Answering](https://mlsanity.com/paper/2004.04906/) (2020)
- [HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering](https://mlsanity.com/paper/1809.09600/) (2018)
- [Not what you've signed up for: Compromising Real-World LLM-Integrated Applications with Indirect Prompt Injection](https://mlsanity.com/paper/2302.12173/) (2023)
- [InjecAgent: Benchmarking Indirect Prompt Injections in Tool-Integrated Large Language Model Agents](https://mlsanity.com/paper/2403.02691/) (2024)
- [Large Dual Encoders Are Generalizable Retrievers](https://mlsanity.com/paper/2112.07899/) (2021)
- [AgentPoison: Red-teaming LLM Agents via Poisoning Memory or Knowledge Bases](https://mlsanity.com/paper/2407.12784/) (2024)
- [Towards Expert-Level Medical Question Answering with Large Language Models](https://mlsanity.com/paper/2305.09617/) (2023)
- [The Faiss library](https://mlsanity.com/paper/2401.08281/) (2024)
- [Formalizing and Benchmarking Prompt Injection Attacks and Defenses](https://mlsanity.com/paper/2310.12815/) (2023)
- [PoisonedRAG: Knowledge Corruption Attacks to Retrieval-Augmented Generation of Large Language Models](https://mlsanity.com/paper/2402.07867/) (2024)
- [The Good and The Bad: Exploring Privacy Issues in Retrieval-Augmented Generation (RAG)](https://mlsanity.com/paper/2402.16893/) (2024)
- [DataSentinel: A Game-Theoretic Detection of Prompt Injection Attacks](https://mlsanity.com/paper/2504.11358/) (2025)
- [Securing AI Agents with Information-Flow Control](https://mlsanity.com/paper/2505.23643/) (2025)
- [Bias and Unfairness in Information Retrieval Systems: New Challenges in the LLM Era](https://mlsanity.com/paper/2404.11457/) (2024)
- [BadRAG: Identifying Vulnerabilities in Retrieval Augmented Generation of Large Language Models](https://mlsanity.com/paper/2406.00083/) (2024)
- [Machine-Made Media: Monitoring the Mobilization of Machine-Generated Articles on Misinformation and Mainstream News Websites](https://mlsanity.com/paper/2305.09820/) (2023)
- [Neural Retrievers are Biased Towards LLM-Generated Content](https://mlsanity.com/paper/2310.20501/) (2023)
- [Topic-FlipRAG: Topic-Orientated Adversarial Opinion Manipulation Attacks to Retrieval-Augmented Generation Models](https://mlsanity.com/paper/2502.01386/) (2025)
- [Real AI Agents with Fake Memories: Fatal Context Manipulation Attacks on Web3 Agents](https://mlsanity.com/paper/2503.16248/) (2025)
- [Backdoor Attacks on Dense Retrieval via Public and Unintentional Triggers](https://mlsanity.com/paper/2402.13532/) (2024)
- [Towards More Robust Retrieval-Augmented Generation: Evaluating RAG Under Adversarial Poisoning Attacks](https://mlsanity.com/paper/2412.16708/) (2024)
- [Invisible Relevance Bias: Text-Image Retrieval Models Prefer AI-Generated Images](https://mlsanity.com/paper/2311.14084/) (2023)
- [WikiCheck: An end-to-end open source Automatic Fact-Checking API based on Wikipedia](https://mlsanity.com/paper/2109.00835/) (2021)
- [How Not to Detect Prompt Injections with an LLM](https://mlsanity.com/paper/2507.05630/) (2025)
- [Bias Amplification in RAG: Poisoning Knowledge Retrieval to Steer LLMs](https://mlsanity.com/paper/2506.11415/) (2025)
- [SMART-RAG: Selection using Determinantal Matrices for Augmented Retrieval](https://mlsanity.com/paper/2409.13992/) (2024)

## Built on by (most cited first, 2 of 2)

- [Personalization Meets Safety:Mechanisms,Risks,and Mitigations in Personalized LLMs](https://mlsanity.com/paper/2606.09038/) (2026)
- [InceptionRAG: Stealthy Poisoning Attack Against Retrieval-Augmented Generation](https://mlsanity.com/paper/2609.16818/) (2026)

## Abstract (arXiv)

When answering user queries, LLMs often retrieve knowledge from external sources stored in retrieval-augmented generation (RAG) databases. These are often populated from unvetted sources, e.g. the open web, and can contain maliciously crafted data. This paper studies attacks that can manipulate the context retrieved by LLMs from such RAG databases. Prior work on such context manipulation primarily injects false or toxic content, which can often be detected by fact-checking or linguistic analysis. A more subtle threat, which we call epistemic bias injection (EBI), is where adversaries inject factually correct yet epistemically biased passages that systematically favor one side of an open-ended issue. Although linguistically coherent and truthful, such adversarial passages effectively crowd out alternative viewpoints during retrieval from the RAG and push LLM outputs towards an attack-desired stance. As a core contribution, we propose a novel characterization of the problem: We give a geometric metric that quantifies stance polarity and epistemic bias. This metric can be computed directly on embeddings of text passages. Leveraging it, we construct EBI attacks and develop a lightweight prototype defense called BiasDef for them. We evaluate them both on a comprehensive benchmark constructed from public question answering datasets. Our results show that: (1) the proposed attack induces significant stance polarity shifts, effectively evading existing retrieval-based sanitization defenses, and (2) BiasDef substantially reduces adversarial retrieval and epistemic bias in LLM's answers. Overall, this demonstrates the new threat as well as the ease of employing epistemic bias metrics for filtering in RAG-enabled LLMs.