Risk-Aware LLM Agents for Geospatial Data Retrieval: Design and Preliminary Adversarial Evaluation
Kyle Gao, Joel Cumming, Jonathan Li, Linlin Xu, David A. Clausi
Why It Matters
What makes this one worth your time
This work is relevant for AI researchers and engineers working on integrating natural language processing with geospatial data systems, offering potential improvements in environmental monitoring and disaster response.
A framework using LLMs for safe and efficient geospatial data retrieval via natural language queries.
Summary
The paper introduces a framework that uses large language models (LLMs) to facilitate the retrieval of geospatial data from cloud-based catalogues through natural language queries. It employs three agents to ensure safe and efficient data access and demonstrates preliminary robustness in adversarial settings.
Key contributions
- Development of a risk-aware LLM-driven framework for geospatial data retrieval.
- Introduction of a multi-agent system for intent interpretation and API call generation.
- Preliminary adversarial evaluation highlighting the need for adaptive defenses.
Notable insights
- The use of a modular framework with interchangeable API schemas for platform portability.
- The integration of a Guardrail agent to enhance safety and policy enforcement in API interactions.
Possible limitations
- Not stated in the abstract
Abstract
arXiv:2606.15077v1 Announce Type: new Abstract: We present an LLM-driven framework for retrieving remote sensing data from cloud-based geospatial catalogues using natural language queries. The system converts user intent into structured API calls, enabling efficient access to satellite imagery and environmental datasets. The architecture integrates three agents: Guardrail for safety and policy enforcement, General-QA for intent interpretation, and Recommender-Analyst for schema-aware API call generation. This coordinated design ensures reliable, semantically aligned interaction with external data services. The modular framework is portable across platforms through API schema substitution and supports applications in environmental monitoring, disaster response, and climate analysis. It establishes a scalable interface between user intent and geospatial infrastructure, enabling streamlined and automated Earth observation workflows. Preliminary experiments under adversarial multi-turn settings show that prompt-level safety instructions improve robustness, although rare high-impact failures persist in API manipulation scenarios and highlight the need for adaptive, system-level defenses that balance safety, usability, and cost efficiency, which motivates the use of our intercept-level Guardrail agent.