Problem
Research questions and friction points this paper is trying to address.
Multi-Agent GIS
Security Risk
Agentic Systems
Prompt Hardening
Adversarial Robustness
Securing Multi-Agent GIS Systems: Risk Evaluation and Prompt Hardening Optimization
🤖 AI Summary
This study addresses the critical security challenges faced by multi-agent GIS systems when performing complex spatial analyses, where existing approaches lack systematic risk assessment and robust defense mechanisms. To enhance system resilience without compromising task performance, this work proposes a security-oriented multi-agent GIS framework that innovatively treats prompts as structured signatures. The framework integrates modular state machines to model agent behavior, employs an adaptive attacker-driven red-teaming approach powered by large language models, incorporates a deterministic binary adjudicator, and leverages adversarial prompt optimization techniques. Experimental results demonstrate that the proposed method effectively identifies and mitigates security vulnerabilities in commercial geospatial systems, significantly improving their defensive capabilities against multi-round adversarial attacks.
📝 Abstract
Agentic systems are increasingly integrated with geographic information systems (GIS), where multi-agent 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 multi-agent 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 multi-turn 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.
Innovation
Methods, ideas, or system contributions that make the work stand out.
multi-agent GIS
prompt hardening
red-teaming
state-machine orchestration
adversarial demonstrations
🔎 Similar Papers
No similar papers found.
