Smart Cities Threat Intelligence Driven by Artificial Intelligence for Enhanced Cyber Resilience Using Federated Learning

Smart cities rely on interconnected technologies, making them vulnerable to sophisticated cyber threats. Integrating Artificial Intelligence (AI) with threat intelligence enhances cyber resilience by enabling proactive security measures. However, existing methods often suffer from centralized data processing, raising privacy concerns and limiting real-time adaptability. To address these issues, this study proposes a Federated Transfer Learning for Adaptive Threat Intelligence (FTL-ATI) framework, which leverages Federated Learning (FL) to ensure decentralized data security while utilizing transfer learning to enhance threat detection efficiency across diverse smart city environments. The proposed framework enables collaborative learning among multiple nodes without sharing raw data, improving threat identification and response times. Experimental results demonstrate that FTL-ATI enhances cyber resilience by achieving higher detection accuracy, reducing false positives, and ensuring real-time adaptability to evolving cyber threats. This approach significantly strengthens smart city cybersecurity, making infrastructures more resilient against emerging cyberattacks while maintaining privacy and efficiency.