AI agents, powered by large language models (LLMs), have transformed human-computer interactions by enabling seamless, natural, and context-aware communication. While these advancements offer immense utility, they also inherit and amplify inherent safety risks such as bias, fairness, hallucinations, privacy breaches, and a lack of transparency. This paper investigates a critical vulnerability: adversarial attacks targeting the LLM core within AI agents. Specifically, we test the hypothesis that a deceptively simple adversarial prefix, such as \textit{Ignore the document}, can compel LLMs to produce dangerous or unintended outputs by bypassing their contextual safeguards. Through experimentation, we demonstrate a high attack success rate (ASR), revealing the fragility of existing LLM defenses. These findings emphasize the urgent need for robust, multi-layered security measures tailored to mitigate vulnerabilities at the LLM level and within broader agent-based architectures.
"Glue pizza and eat rocks" -- Exploiting Vulnerabilities in Retrieval-Augmented Generative Models
Arizona State University(亚利桑那州立大学);Michigan State University(密歇根州立大学);University of North Carolina at Chapel Hill(北卡罗来纳大学教堂山分校);University of Virginia(弗吉尼亚大学)
VisualTrap: A Stealthy Backdoor Attack on GUI Agents via Visual Grounding Manipulation
University of Science and Technology of China(中国科学技术大学);National University of Singapore(新加坡国立大学);East China University of Science and Technology(东华大学)
DCMI: A Differential Calibration Membership Inference Attack Against Retrieval-Augmented Generation
School of Cyber Science and Technology, Shandong University(山东大学计算机科学与技术学院);Department of Engineering Software, School of Civil Engineering, Shandong University(山东大学土木工程学院工程软件系);State Key Laboratory of Cryptography and Digital Economy Security, Shandong University(山东大学密码与数字经济安全国家重点实验室);Shandong Key Laboratory of Artificial Intelligence Security, Shandong University(山东省人工智能安全重点实验室)
BioProVLA-Agent: An Affordable, Protocol-Driven, Vision-Enhanced VLA-Enabled Embodied Multi-Agent System with Closed-Loop-Capable Reasoning for Biological Laboratory Manipulation
Key Laboratory of Smart Manufacturing in Energy Chemical Process Ministry of Education, East China University of Science and Technology, Shanghai, CN(能源化工过程智能制造教育部重点实验室,东华大学,上海,中国);Department of Computer Science and Engineering, East China University of Science and Technology, Shanghai, CN(东华大学计算机科学与工程系,上海,中国);Department of Laboratory Medicine, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, CN(复旦大学附属瑞金医院检验医学科,上海,中国);School of Information Science and Technology, Shihezi University, Shihezi, CN(石河子大学信息科学与技术学院,石河子,中国)
RoBCtrl: Attacking GNN-Based Social Bot Detectors via Reinforced Manipulation of Bots Control Interaction
University of Science and Technology of China(科学技术大学);Beihang University(北京航空航天大学);National University of Singapore(国立新加坡大学);Peking University(北京大学);University of Illinois Chicago(伊利诺伊大学香槟分校);Institute for Clarity in Documentation(文档清晰研究所);Inria Paris-Rocquencourt(巴黎-拉德克利夫研究所);Rajiv Gandhi University(拉贾·甘地大学);Tsinghua University(清华大学);Palmer Research Laboratories(帕尔默研究实验室)
A Taxonomy of Pix Fraud in Brazil: Attack Methodologies, AI-Driven Amplification, and Defensive Strategies
Graduate Program in Software Engineering (PPGES) Advanced Computing Studies Laboratory (LEA) \& AI Horizon Labs Federal University of Pampa (UNIPAMPA), Alegrete, Brazil