phenomenon · first seen 5 Oct 2026
retrieval-augmented generation poisoning
Also: RAG poisoning, knowledge poisoning, knowledge poisoning attack, retrieval poisoning attack, adversarial attacks on retrieval-augmented generation, knowledge poisoning attack on retrieval-augmented generation
Attacks that inject malicious documents into a retrieval corpus so generated answers are corrupted.
Mentions per week
PapersCodeDiscussion
Earliest sources
- Bounded Provisional Visibility: Controlling Poisoning Exposure in Continuously Ingested RAG Vector StoresPapers · 5 Oct 2026 · Chuhong Xu, Lu Yi, Gangzhen Qian, Kainan Zhou, Zhaoyi Li, Hang Xiao
- GraphRAG-LTM: A Layered Threat Modeling Framework for Attacks on Graph Retrieval-Augmented GenerationPapers · 14 Oct 2026 · Yining Tang, Junyu Zhang, Haomiao Qiu, Yongchao Liu, Jiajun Zheng, Zhao Wang, Zi…
- Learning When to Trust Medical Knowledge: Regression-Safe Routing for Clinical ReasoningPapers · 14 Oct 2026 · Hui Zhang, Yujia Zeng, Yuhao Xin
- Ranking Is Not Deciding: Benchmark Shortcuts and Operating-Point Collapse in Query-Time RAG Poisoning DefensesPapers · 24 Oct 2026 · Pushkal Kumar, Arvin Sahaym
- Factual Yet One-Sided: Stance Poisoning and Evidence Balancing in Retrieval-Augmented GenerationPapers · 25 Oct 2026 · Puyuan Yang, Shaoxuan Dong, Liujie Wu, Xiaochun Cheng, Bo Li
- Prevented is Not Protected: What Attack Success Rate Hides about RAGPapers · 29 Oct 2026 · Chahana Dahal, Zuobin Xiong
Latest sources
- Security of Deep Research Agents: Attack Surface, Defenses, Benchmarks, and Agenda Seeding against Research PlanningPapers · 15 Nov 2026 · Yang Luo, Guangpu Chen, Yichen Liu, Zhaoyu Chen, Yifeng Gao, Zhiyang Ding, Qingn…
- One Fabrication Is Enough: A Fake News Article Changes the Forecasts of LLM Forecasting SystemsPapers · 11 Nov 2026 · Zeyu Zhang, Bradly C. Stadie
- IAMRAGBENCH: Benchmarking Authorization Expansions in LLM-Generated AWS IAM Policies under RAG PoisoningPapers · 2 Nov 2026 · Hasan Kaplan, Indika Kumara, Willem-Jan van den Heuvel, Damian Andrew Tamburri
- Illusory Truth or Salient Anomaly? Model-Dependent Responses to Repeated Poisoned Passages in RAG-Style ContextsPapers · 29 Oct 2026 · Mehrdad Ghassabi, Audrina Ebrahimi, Sadra Hakim, Hamidreza Baradaran Kashani, Za…
- Prevented is Not Protected: What Attack Success Rate Hides about RAGPapers · 29 Oct 2026 · Chahana Dahal, Zuobin Xiong
- Factual Yet One-Sided: Stance Poisoning and Evidence Balancing in Retrieval-Augmented GenerationPapers · 25 Oct 2026 · Puyuan Yang, Shaoxuan Dong, Liujie Wu, Xiaochun Cheng, Bo Li
- Ranking Is Not Deciding: Benchmark Shortcuts and Operating-Point Collapse in Query-Time RAG Poisoning DefensesPapers · 24 Oct 2026 · Pushkal Kumar, Arvin Sahaym
- Learning When to Trust Medical Knowledge: Regression-Safe Routing for Clinical ReasoningPapers · 14 Oct 2026 · Hui Zhang, Yujia Zeng, Yuhao Xin
- GraphRAG-LTM: A Layered Threat Modeling Framework for Attacks on Graph Retrieval-Augmented GenerationPapers · 14 Oct 2026 · Yining Tang, Junyu Zhang, Haomiao Qiu, Yongchao Liu, Jiajun Zheng, Zhao Wang, Zi…
- Bounded Provisional Visibility: Controlling Poisoning Exposure in Continuously Ingested RAG Vector StoresPapers · 5 Oct 2026 · Chuhong Xu, Lu Yi, Gangzhen Qian, Kainan Zhou, Zhaoyi Li, Hang Xiao