DiscourseFlip: An Oblique Discourse-Level Opinion Manipulation Attack against Black-box Retrieval-Augmented Generation
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866916071872659456 |
|---|---|
| author | Gong, Yuyang Chen, Miaokun Liu, Jiawei Chen, Zhuo He, Guoxiu Lu, Wei Wang, XiaoFeng Liu, Xiaozhong |
| author_facet | Gong, Yuyang Chen, Miaokun Liu, Jiawei Chen, Zhuo He, Guoxiu Lu, Wei Wang, XiaoFeng Liu, Xiaozhong |
| contents | Retrieval-Augmented Generation (RAG) systems are widely deployed and increasingly influential, but their reliance on external corpora exposes new security risks from poisoned retrieval content. Existing RAG attacks are largely focusing on individual queries or narrow topic-local query sets, which limits their practical reach and offers limited camouflage in real-world settings. In this paper, we introduce discourse-level opinion manipulation, a new threat model in which coordinated influence across a semantic query network induces opinion shifts over a holistic, multi-topic query space. We formalize this threat in a black-box setting and propose DiscourseFlip, an agentic, graph-guided attack that dynamically allocates a limited poisoning budget to maximize discourse-level opinion deviation. Extensive experiments demonstrate that DiscourseFlip consistently induces targeted opinion shifts across the contextualized query network and significantly outperforms existing baselines in terms of coverage and effectiveness. User studies further confirm that DiscourseFlip is effective while remaining well camouflaged from user detection. Moreover, systematic analyses show that existing mitigation strategies are ineffective against discourse-level manipulation, underscoring the urgent need for more robust and adaptive defenses to address discourse-level vulnerabilities. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2606_01212 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | DiscourseFlip: An Oblique Discourse-Level Opinion Manipulation Attack against Black-box Retrieval-Augmented Generation Gong, Yuyang Chen, Miaokun Liu, Jiawei Chen, Zhuo He, Guoxiu Lu, Wei Wang, XiaoFeng Liu, Xiaozhong Computation and Language Artificial Intelligence Cryptography and Security Information Retrieval Retrieval-Augmented Generation (RAG) systems are widely deployed and increasingly influential, but their reliance on external corpora exposes new security risks from poisoned retrieval content. Existing RAG attacks are largely focusing on individual queries or narrow topic-local query sets, which limits their practical reach and offers limited camouflage in real-world settings. In this paper, we introduce discourse-level opinion manipulation, a new threat model in which coordinated influence across a semantic query network induces opinion shifts over a holistic, multi-topic query space. We formalize this threat in a black-box setting and propose DiscourseFlip, an agentic, graph-guided attack that dynamically allocates a limited poisoning budget to maximize discourse-level opinion deviation. Extensive experiments demonstrate that DiscourseFlip consistently induces targeted opinion shifts across the contextualized query network and significantly outperforms existing baselines in terms of coverage and effectiveness. User studies further confirm that DiscourseFlip is effective while remaining well camouflaged from user detection. Moreover, systematic analyses show that existing mitigation strategies are ineffective against discourse-level manipulation, underscoring the urgent need for more robust and adaptive defenses to address discourse-level vulnerabilities. |
| title | DiscourseFlip: An Oblique Discourse-Level Opinion Manipulation Attack against Black-box Retrieval-Augmented Generation |
| topic | Computation and Language Artificial Intelligence Cryptography and Security Information Retrieval |
| url | https://arxiv.org/abs/2606.01212 |