CPA-RAG:Covert Poisoning Attacks on Retrieval-Augmented Generation in Large Language Models
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866912395269505024 |
|---|---|
| author | Li, Chunyang Zhang, Junwei Cheng, Anda Ma, Zhuo Li, Xinghua Ma, Jianfeng |
| author_facet | Li, Chunyang Zhang, Junwei Cheng, Anda Ma, Zhuo Li, Xinghua Ma, Jianfeng |
| contents | Retrieval-Augmented Generation (RAG) enhances large language models (LLMs) by incorporating external knowledge, but its openness introduces vulnerabilities that can be exploited by poisoning attacks. Existing poisoning methods for RAG systems have limitations, such as poor generalization and lack of fluency in adversarial texts. In this paper, we propose CPA-RAG, a black-box adversarial framework that generates query-relevant texts capable of manipulating the retrieval process to induce target answers. The proposed method integrates prompt-based text generation, cross-guided optimization through multiple LLMs, and retriever-based scoring to construct high-quality adversarial samples. We conduct extensive experiments across multiple datasets and LLMs to evaluate its effectiveness. Results show that the framework achieves over 90\% attack success when the top-k retrieval setting is 5, matching white-box performance, and maintains a consistent advantage of approximately 5 percentage points across different top-k values. It also outperforms existing black-box baselines by 14.5 percentage points under various defense strategies. Furthermore, our method successfully compromises a commercial RAG system deployed on Alibaba's BaiLian platform, demonstrating its practical threat in real-world applications. These findings underscore the need for more robust and secure RAG frameworks to defend against poisoning attacks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_19864 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | CPA-RAG:Covert Poisoning Attacks on Retrieval-Augmented Generation in Large Language Models Li, Chunyang Zhang, Junwei Cheng, Anda Ma, Zhuo Li, Xinghua Ma, Jianfeng Cryptography and Security Retrieval-Augmented Generation (RAG) enhances large language models (LLMs) by incorporating external knowledge, but its openness introduces vulnerabilities that can be exploited by poisoning attacks. Existing poisoning methods for RAG systems have limitations, such as poor generalization and lack of fluency in adversarial texts. In this paper, we propose CPA-RAG, a black-box adversarial framework that generates query-relevant texts capable of manipulating the retrieval process to induce target answers. The proposed method integrates prompt-based text generation, cross-guided optimization through multiple LLMs, and retriever-based scoring to construct high-quality adversarial samples. We conduct extensive experiments across multiple datasets and LLMs to evaluate its effectiveness. Results show that the framework achieves over 90\% attack success when the top-k retrieval setting is 5, matching white-box performance, and maintains a consistent advantage of approximately 5 percentage points across different top-k values. It also outperforms existing black-box baselines by 14.5 percentage points under various defense strategies. Furthermore, our method successfully compromises a commercial RAG system deployed on Alibaba's BaiLian platform, demonstrating its practical threat in real-world applications. These findings underscore the need for more robust and secure RAG frameworks to defend against poisoning attacks. |
| title | CPA-RAG:Covert Poisoning Attacks on Retrieval-Augmented Generation in Large Language Models |
| topic | Cryptography and Security |
| url | https://arxiv.org/abs/2505.19864 |