一句话：首个针对 LLM Agent 应用漏洞的系统测量研究，归纳 14 类漏洞、16 种根因并评估缓解策略的权衡。
随着大语言模型在内容理解与问答上的优势，LLM Agent 应用迅速普及，开发者广泛构建此类应用以简化流程、提升效率。然而这些应用已被发现存在大量严重漏洞与安全问题，因此亟需系统理解其漏洞特性与缓解方式。本文首次对 Agent 应用的漏洞、开发者的缓解实践及其挑战与权衡展开综合研究：基于 221 个真实漏洞的分析，覆盖 7 个组件，识别出 14 类漏洞与 16 种根因；进而考察开发者的应对行为，评估多种缓解策略的有效性，并探讨漏洞缓解中的实际困难与不可避免的权衡。最终提炼出 12 条关键发现，讨论其对 Agent 应用开发者、维护者与安全研究者的启示，并给出未来研究方向建议。
The advantages of large language models (LLMs) in content comprehension and question answering have led to the rapid emergence of LLM agent. Developers across diverse domains are actively building their own agent applications (apps), as these apps can streamline workflows, boost efficiency, or deliver innovative solutions, thereby enhancing the competitiveness of their products. Agent apps are playing an increasingly important role in our daily lives. However, numerous serious vulnerabilities and security issues have been identified in these apps. To effectively manage future security risks, it is essential to systematically understand the unique characteristics of agent app vulnerabilities and their mitigation. In this paper, we present the first comprehensive study on the vulnerabilities of agent apps, the mitigation practices of app developers, and the associated challenges and trade-offs. We identify 14 types of vulnerabilities and 16 root causes across 7 components, based on an analysis of 221 real-world vulnerabilities. Our study further investigates developer reactions, evaluates the effectiveness of various mitigation strategies, and explores the practical challenges and inevitable trade-offs in vulnerability mitigation. Finally, we distill 12 key findings, discuss their implications for agent app developers, maintainers, and security researchers, and offer suggestions for future research directions.
DOI 原文 ·
@inproceedings{paperbot9394,
title = {Security Debt in LLM Agent Applications: A Measurement Study of Vulnerabilities and Mitigation Trade-offs},
author = {Zhuoxiang Shen and Jiarun Dai and Yuan Zhang and Min Yang},
booktitle = {ASE},
year = {2025},
doi = {10.1109/ase63991.2025.00053}
}