Vis-Poison is a novel attack that poisons multimodal retrieval-augmented generation systems by inserting attacker-controlled images as visual evidence, without altering any textual metadata. The attack uses an automated multi-agent approach to create visually plausible poisoned images and has been tested on two multimodal RAG pipelines, four embedding models, and six generation models. In black-box settings, Vis-Poison achieves an end-to-end success rate between 40.16% and 65.40% against 30,000-entry knowledge bases, and remains effective against various multimodal large language models with an average success rate above 60%.
By Rujin Liang, Zhongpu Chen, Yuhao Lei, Xin Miao
Multimodal Retrieval Augmented Generation (M-RAG) is increasingly vulnerable to adversarial attacks where malicious data are crafted to produce embeddings that align with benign entries in the vector...
arXiv:2511.18921v2 Announce Type: replace
Abstract: Backdoor attacks undermine the reliability and trustworthiness of machine learning systems by injecting hidden behaviors that can be maliciously ac...
By Juncheng Li, Yige Li, Hanxun Huang, Yunhao Chen, Xin Wang, Yixu Wang, Xingjun Ma, Yu-Gang Jiang
arXiv:2606. 18310v1 Announce Type: cross Abstract: Injecting malicious knowledge into retrieval-augmented generation (RAG) systems can manipulate retrieved evidence and mislead downstream generation, posing a serious security threat for AI applications.
By Xinru Liu, Xianglong Zhang, Di Cai, Zhumin Chen, Pengfei Hu, Xin Xin
arXiv:2608. 28389v1 Announce Type: cross Abstract: Retrieval-augmented generation (RAG) augments LLMs with external documents, but public or user-editable sources expose RAG systems to data poisoning: attackers can inject malicious documents to steer outputs toward targeted answers.
By Jaewon Jung, Haizhong Zheng, Hongsun Jang, Jaeyong Song, Beidi Chen, Jinho Lee
Repeat-After-Me is a black-box adaptive visual prompt injection technique that can reveal personally identifiable information or trigger malicious tool calls in both open-weight and commercial vision‑language models, achieving attack success rates above 80% on Qwen3.6‑27B and 47% on GPT‑5.5. The method works even when the benign user prompt is unrelated to the injected task and does not explicitly authorize it, and it retains significant effectiveness when transferred across models or optimized on surrogate systems. In a real‑world OpenClaw Discord deployment, a minimally injected image can overwrite TOOLS.md, enabling remote code execution and secret exfiltration.
By Sizhe Chen, Yu-Lin Tsai, Ivan Evtimov, Kamalika Chaudhuri, Raluca Ada Popa, David Wagner, Arman Zharmagambetov
arXiv:2406.00083v3 Announce Type: replace-cross
Abstract: Retrieval-Augmented Generation (RAG) enhances Large Language Models (LLMs) by retrieving relevant information from external knowledge bases t...
By Jiaqi Xue, Mengxin Zheng, Yebowen Hu, Fei Liu, Xun Chen, Qian Lou
The paper introduces SHAQ, a defense called Shadow Query Generation that protects document embeddings in vector databases from embedding inversion attacks. SHAQ replaces direct embeddings with diverse shadow queries generated by a language model, thereby decomposing document semantics and decoupling stored embeddings from the original text. Experiments on various IR datasets show that SHAQ significantly lowers recovery rates, defends more tokens than baseline methods, and even improves retrieval utility.
By Xinguo Feng, Zhongkui Ma, Zihan Wang, Chuan Yan, Guowei Yang, Alsharif Abuadbba, Guangdong Bai
The paper introduces COMIC, a reference‑aware safety gate designed for multimodal large language models (MLLMs). COMIC detects the operation requested by a user, identifies visual targets through OCR and open‑vocabulary proposals, and evaluates safety on explicit operation‑target pairs, using max‑risk aggregation and quality‑aware routing to decide whether to allow or block a request. Experiments on several open‑source MLLMs and jailbreak benchmarks show that COMIC improves robustness while maintaining benign utility and efficiency.
By Md Abdullahil Oaphy, Anhao Xiang, Zongxing Xie, Huayue Gu, Chenyu Wang, Honghui Xu
arXiv:2603. 29418v2 Announce Type: replace-cross Abstract: Although multimodal large language models (MLLMs) are increasingly deployed in real-world applications, their instruction-following behavior leaves them vulnerable to prompt injection attacks.
By Meiwen Ding, Song Xia, Chenqi Kong, Xudong Jiang
arXiv:2507. 07947v4 Announce Type: replace-cross Abstract: Recent advances in generative models, such as diffusion models, have raised concerns related to privacy, copyright infringement, and data stewardship.
By Sol Yarkoni, Mahmood Sharif, Roi Livni
arXiv:2606. 02643v1 Announce Type: cross Abstract: Retrieval-Augmented Generation (RAG)-enhanced LLM systems, while powerful, introduce substantial inference costs due to the inclusion of an extra multi-stage pipeline that dynamically retrieves and synthesizes information from external knowledge sources.
By Chengliang Liu, Liangbo Ning, Yujuan Ding, Wenqi Fan