AI-driven cyber-resilience framework for smart city infrastructure: integrating IoT, predictive analytics and digital twin governance

Abstract

The rapid digitalization of urban infrastructure has transformed smart cities into complex cyber-physical ecosystems where Internet of Things (IoT) devices, cloud platforms, and artificial intelligence (AI) systems operate as critical enablers of public services. However, this interconnected architecture expands the attack surface of essential assets such as bridges, transport networks, energy grids, and water systems. Cyber incidents targeting sensor networks or data management platforms can cascade into physical disruptions, undermining public safety and trust. This paper proposes an AI-driven cyber-resilience framework designed specifically for smart city critical infrastructure. Methodologically, the paper develops a layered architecture model and evaluates its resilience capacity against representative threat scenarios, including data spoofing, denial-of-service attacks, and sensor manipulation. The analysis demonstrates that integrating predictive analytics within a digital twin governance structure significantly improves early detection rates while reducing false positives and response latency. Beyond technical contributions, the study outlines policy implications for municipal administrations, emphasizing interoperability, data protection compliance, and cross-sector coordination. The proposed framework advances the transition from reactive cybersecurity strategies to anticipatory cyber-resilience in intelligent urban systems.