Taxonomy, Evaluation and Exploitation of IPI-Centric LLM Agent Defense Frameworks
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909913425379328 |
|---|---|
| author | Ji, Zimo Wang, Xunguang Li, Zongjie Ma, Pingchuan Gao, Yudong Wu, Daoyuan Yan, Xincheng Tian, Tian Wang, Shuai |
| author_facet | Ji, Zimo Wang, Xunguang Li, Zongjie Ma, Pingchuan Gao, Yudong Wu, Daoyuan Yan, Xincheng Tian, Tian Wang, Shuai |
| contents | Large Language Model (LLM)-based agents with function-calling capabilities are increasingly deployed, but remain vulnerable to Indirect Prompt Injection (IPI) attacks that hijack their tool calls. In response, numerous IPI-centric defense frameworks have emerged. However, these defenses are fragmented, lacking a unified taxonomy and comprehensive evaluation. In this Systematization of Knowledge (SoK), we present the first comprehensive analysis of IPI-centric defense frameworks. We introduce a comprehensive taxonomy of these defenses, classifying them along five dimensions. We then thoroughly assess the security and usability of representative defense frameworks. Through analysis of defensive failures in the assessment, we identify six root causes of defense circumvention. Based on these findings, we design three novel adaptive attacks that significantly improve attack success rates targeting specific frameworks, demonstrating the severity of the flaws in these defenses. Our paper provides a foundation and critical insights for the future development of more secure and usable IPI-centric agent defense frameworks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_15203 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Taxonomy, Evaluation and Exploitation of IPI-Centric LLM Agent Defense Frameworks Ji, Zimo Wang, Xunguang Li, Zongjie Ma, Pingchuan Gao, Yudong Wu, Daoyuan Yan, Xincheng Tian, Tian Wang, Shuai Cryptography and Security Artificial Intelligence Large Language Model (LLM)-based agents with function-calling capabilities are increasingly deployed, but remain vulnerable to Indirect Prompt Injection (IPI) attacks that hijack their tool calls. In response, numerous IPI-centric defense frameworks have emerged. However, these defenses are fragmented, lacking a unified taxonomy and comprehensive evaluation. In this Systematization of Knowledge (SoK), we present the first comprehensive analysis of IPI-centric defense frameworks. We introduce a comprehensive taxonomy of these defenses, classifying them along five dimensions. We then thoroughly assess the security and usability of representative defense frameworks. Through analysis of defensive failures in the assessment, we identify six root causes of defense circumvention. Based on these findings, we design three novel adaptive attacks that significantly improve attack success rates targeting specific frameworks, demonstrating the severity of the flaws in these defenses. Our paper provides a foundation and critical insights for the future development of more secure and usable IPI-centric agent defense frameworks. |
| title | Taxonomy, Evaluation and Exploitation of IPI-Centric LLM Agent Defense Frameworks |
| topic | Cryptography and Security Artificial Intelligence |
| url | https://arxiv.org/abs/2511.15203 |