A METHODOLOGICAL FRAMEWORK FOR ASSESSING THE ROLE OF ARTIFICIAL INTELLIGENCE IN THE CYBERSECURITY OF MODERN GIS SYSTEMS

Abstract

This paper proposes an integrated methodological framework for assessing the role of artificial intelligence in the cybersecurity of contemporary geographic information systems (GIS). The framework is structured around three interrelated dimensions: spatial data quality, the resilience of anomaly detection models to geolocation manipulation, and the security architecture of geospatial service environments. The study aims to provide a coherent, simulation-based foundation for systematically examining the relationships among these dimensions in a controlled and reproducible research setting. The framework is demonstrated using synthetic yet operationally realistic data representing an urban public transport environment. The methodological design combines: (i) formal assessment of spatial data quality in accordance with ISO 19157-1:2023 and ISO 19115-1:2014; (ii) comparative evaluation of anomaly detection models implemented in single-channel and multimodal configurations; and (iii) assessment of OGC API–based geospatial services under conventional trust-based and Zero Trust security architectures. The simulation environment includes controlled data quality degradation, GNSS spoofing scenarios, and selected API abuse patterns. The results indicate that the proposed framework provides a coherent basis for analysing the effects of data quality on anomaly detection, the response of multimodal AI models to geolocation manipulation. Under simulated disruption conditions, multimodal analytical configurations and Zero Trust architectures showed greater robustness than conventional approaches. The framework offers a methodological foundation for future research on AI-supported GIS cybersecurity.