English(EN) CiteShade: Citation Laundering in Multi-Source Retrieval-Augmented Generation and Its Counterfactual Defense

新的“CiteShade”攻击利用RAG系统中的引用洗白

作者 PulseAugur 编辑部 · [1 个来源] · 2026-09-15 04:00

一篇新论文介绍了“CiteShade”，一种针对检索增强生成（RAG）系统的新型攻击向量。该攻击允许控制单个数据源的对手操纵语言模型生成不正确答案，并将其错误地归因于受信任的来源。研究表明，这种引用洗白会显著增加多跳问答场景中的错误答案率，凸显了模型处理引用的一个漏洞。 AI

影响凸显了RAG系统中的一个新漏洞，可能影响AI生成信息的可靠性，并需要新的防御机制。

排序理由该集群包含一篇详细介绍AI系统新攻击向量的研究论文。[lever_c_demoted from research: ic=1 ai=1.0]

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 16 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该集群包含一篇详细介绍AI系统新攻击向量的研究论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, safety; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

arXiv cs.CL TIER_1 English(EN) · Guo Fuzheng · 2026-09-15 04:00

CiteShade：多源检索增强生成中的引用洗白及其反事实防御

arXiv:2609.15660v1 Announce Type: cross Abstract: Retrieval-augmented generation (RAG) grounds a language model's answers on retrieved external knowledge and returns each answer with citations that identify its sources. Those citations are the user's audit trail: they let a reade…