CROSS-DOMAIN RISK PROPAGATION IN INTERCONNECTED INFRASTRUCTURE ECOSYSTEMS

Modern infrastructure systems are increasingly interdependent across physical, cyber, and logical domains. This interconnectedness, while enhancing operational efficiency, introduces systemic vulnerabilities where failure in one domain can propagate and destabilize others. This paper proposes a hybrid AI-assisted framework to simulate, analyze, and mitigate cross-domain risk propagation. By integrating graph theory, agent-based modeling, and machine learning (including LSTM, GNN, and reinforcement learning), the research provides predictive capabilities and decision support for complex infrastructure ecosystems. Case studies of real-world cascading failures such as the 2003 North American blackout and the 2021 Texas winter storm validate the frameworks relevance. The work contributes a practical toolset and metrics for policymakers and infrastructure operators seeking to design more resilient, adaptive systems.