English(EN) DiscourseFlip: An Oblique Discourse-Level Opinion Manipulation Attack against Black-box Retrieval-Augmented Generation

新的攻击方法操纵跨查询网络的RAG系统意见

作者 PulseAugur 编辑部 · [2 个来源] · 2026-05-31 13:03

研究人员开发了DiscourseFlip，一种针对检索增强生成（RAG）系统的新型攻击方法。该攻击操纵跨相关查询网络的意见，超越了单查询攻击，引发更广泛的转变。实验表明，DiscourseFlip在改变意见方面是有效的，并且保持了良好的隐蔽性，而现有防御措施不足。 AI

影响突显了RAG系统的新漏洞，需要改进防御措施以应对复杂的多主题操纵。

排序理由该集群包含一篇详细介绍针对AI系统的新攻击方法的论文。

AI 生成摘要 · Google Gemini · 来自 2 个来源。我们如何撰写摘要 →

报道来源 [2]

arXiv cs.AI TIER_1 English(EN) · Yuyang Gong, Miaokun Chen, Jiawei Liu, Zhuo Chen, Guoxiu He, Wei Lu, XiaoFeng Wang, Xiaozhong Liu · 2026-06-02 04:00

DiscourseFlip: An Oblique Discourse-Level Opinion Manipulation Attack against Black-box Retrieval-Augmented Generation

arXiv:2606.01212v1 Announce Type: cross Abstract: 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 foc…
arXiv cs.IR (Information Retrieval) TIER_1 English(EN) · Xiaozhong Liu · 2026-05-31 13:03

DiscourseFlip: An Oblique Discourse-Level Opinion Manipulation Attack against Black-box Retrieval-Augmented Generation

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 …