Towards Continuous Security Assessment: Integrating Model-Based Risk Assessment and Large Language Models

Cyber threats and attacks on aircraft ground infrastructure as well as communication, navigation, and surveillance service providers have increased significantly in recent years. Due to the dynamic nature of cyber threats, a once securely designed architecture will not stay secure throughout its life-cycle. Ensuring the safety and security of large systems of systems requires continuous reassessment of security risks. Currently, while useful building blocks exist, no comprehensive framework supports the continuous and automated risk reassessment in aviation. To address this gap, we propose a conceptual framework that integrates model-based system specification, model-based security assessment, and Large Language Models (LLMs) to automate the continuous reassessment of residual risks in cyber-physical systems. This framework dynamically analyzes and adapts to emerging threats, enabling proactive and up-to-date risk management. This paper outlines the high-level requirements of the proposed framework and presents a state-of-the-art analysis of technology readiness and existing gaps necessary for its implementation. To demonstrate its feasibility and effectiveness, we apply the framework to an Air Traffic Management use case.