LogicPoison: Logical Attacks on Graph Retrieval-Augmented Generation

Yilin Xiao, Jin Chen, Qinggang Zhang, Yujing Zhang, Chuang Zhou, Longhao Yang, Lingfei Ren, Xin Yang, Xiao Huang

Abstract

Graph-based Retrieval-Augmented Generation (GraphRAG) enhances the reasoning capabilities of Large Language Models (LLMs) by grounding their responses in structured knowledge graphs. Leveraging community detection and relation filtering techniques, GraphRAG systems demonstrate inherent resistance to traditional RAG attacks, such as text poisoning and prompt injection. However, in this paper, we find that the security of GraphRAG systems fundamentally relies on the topological integrity of the un

Categories

Cite This Resource

@article{llmsec202600154,
  title = {LogicPoison: Logical Attacks on Graph Retrieval-Augmented Generation},
  author = {Yilin Xiao and Jin Chen and Qinggang Zhang and Yujing Zhang and Chuang Zhou and Longhao Yang and Lingfei Ren and Xin Yang and Xiao Huang},
  year = {2026},
  url = {https://arxiv.org/abs/2604.02954},
}