跳到主要导航 跳到搜索 跳到主要内容

Securing Multiagent GIS Systems: Risk Evaluation and Prompt Hardening Optimization

  • Kyle Gao*
  • , Pranavi Kotta
  • , Linlin Lincoln Xu
  • , Jonathan Li
  • , David A. Clausi
  • *此作品的通讯作者
  • Aalto University
  • University of Waterloo
  • University of Calgary

科研成果: 期刊稿件 › 文章 › 同行评审

摘要

Agentic systems are increasingly integrated with geographic information systems (GISs), where multiagent coordination enables complex conversational and spatial analysis but introduces security risks. This work presents a security-oriented framework for risk identification, evaluation, and mitigation in a multiagent GIS system while maintaining adaptability to broader agentic architectures. We test the agentic system of a commercial geospatial partner while developing a modular state-machine–based orchestration framework that abstracts agent behavior into reusable components. We evaluate robustness using a red-teaming framework with an adaptive attacker LLM and a deterministic judge that produces binary outcomes with supporting rationales across multiturn attacks. We further improve resilience with a prompt optimization framework that treats prompts as structured signatures and injects adversarial demonstrations, enabling systematic security improvements without degrading task performance.

源语言英语
期刊论文编号8002505
期刊IEEE Geoscience and Remote Sensing Letters
卷23
DOI
出版状态已出版 - 2026
已对外发布是

学术指纹

探究 'Securing Multiagent GIS Systems: Risk Evaluation and Prompt Hardening Optimization' 的科研主题。它们共同构成独一无二的学术指纹。

引用此