@inproceedings{yu-sun-2026-coverpoison,
title = "{C}over{P}oison: One-to-Many Knowledge Corruption Attacks on Strictly Black-Box {RAG} Systems",
author = "Yu, Zihao and
Sun, Jun",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.377/",
pages = "7966--7981",
ISBN = "979-8-89176-462-0",
abstract = "Retrieval-Augmented Generation (RAG) systems are vulnerable to knowledge corruption attacks, where adversaries inject malicious texts into external corpora to manipulate generated responses. Existing one-to-many attacks can compromise multiple queries with a small number of injected texts, but they typically require white-box access to the target retriever for gradient-based optimization. In contrast, existing black-box attacks are more practical but usually allow each malicious text to target only one query. In this paper, we study one-to-many knowledge corruption attacks against strictly black-box RAG systems, where the attacker has no access to the target retriever, generator, or corpus contents. We propose CoverPoison, a black-box method that combines competitive similarity clustering, coverage-oriented optimization, and T5-based text refinement. Competitive similarity clustering improves query grouping quality, while the coverage-oriented objective encourages each candidate to cover more queries through query-specific retrieval thresholds. Experiments on three datasets show that CoverPoison substantially improves black-box attack effectiveness. Further analyses across RAG configurations, attack objectives, and defenses validate the effectiveness of the proposed method."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="yu-sun-2026-coverpoison">
<titleInfo>
<title>CoverPoison: One-to-Many Knowledge Corruption Attacks on Strictly Black-Box RAG Systems</title>
</titleInfo>
<name type="personal">
<namePart type="given">Zihao</namePart>
<namePart type="family">Yu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jun</namePart>
<namePart type="family">Sun</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>Retrieval-Augmented Generation (RAG) systems are vulnerable to knowledge corruption attacks, where adversaries inject malicious texts into external corpora to manipulate generated responses. Existing one-to-many attacks can compromise multiple queries with a small number of injected texts, but they typically require white-box access to the target retriever for gradient-based optimization. In contrast, existing black-box attacks are more practical but usually allow each malicious text to target only one query. In this paper, we study one-to-many knowledge corruption attacks against strictly black-box RAG systems, where the attacker has no access to the target retriever, generator, or corpus contents. We propose CoverPoison, a black-box method that combines competitive similarity clustering, coverage-oriented optimization, and T5-based text refinement. Competitive similarity clustering improves query grouping quality, while the coverage-oriented objective encourages each candidate to cover more queries through query-specific retrieval thresholds. Experiments on three datasets show that CoverPoison substantially improves black-box attack effectiveness. Further analyses across RAG configurations, attack objectives, and defenses validate the effectiveness of the proposed method.</abstract>
<identifier type="citekey">yu-sun-2026-coverpoison</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.377/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>7966</start>
<end>7981</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T CoverPoison: One-to-Many Knowledge Corruption Attacks on Strictly Black-Box RAG Systems
%A Yu, Zihao
%A Sun, Jun
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F yu-sun-2026-coverpoison
%X Retrieval-Augmented Generation (RAG) systems are vulnerable to knowledge corruption attacks, where adversaries inject malicious texts into external corpora to manipulate generated responses. Existing one-to-many attacks can compromise multiple queries with a small number of injected texts, but they typically require white-box access to the target retriever for gradient-based optimization. In contrast, existing black-box attacks are more practical but usually allow each malicious text to target only one query. In this paper, we study one-to-many knowledge corruption attacks against strictly black-box RAG systems, where the attacker has no access to the target retriever, generator, or corpus contents. We propose CoverPoison, a black-box method that combines competitive similarity clustering, coverage-oriented optimization, and T5-based text refinement. Competitive similarity clustering improves query grouping quality, while the coverage-oriented objective encourages each candidate to cover more queries through query-specific retrieval thresholds. Experiments on three datasets show that CoverPoison substantially improves black-box attack effectiveness. Further analyses across RAG configurations, attack objectives, and defenses validate the effectiveness of the proposed method.
%U https://aclanthology.org/2026.findings-emnlp.377/
%P 7966-7981
Markdown (Informal)
[CoverPoison: One-to-Many Knowledge Corruption Attacks on Strictly Black-Box RAG Systems](https://aclanthology.org/2026.findings-emnlp.377/) (Yu & Sun, Findings 2026)
ACL