Explainable AI and machine learning for robust cybersecurity in smart cities

Abstract

An emerging application of such new technologies is in urban development, with cities increasingly utilizing them to address social, environmental, and urban issues. IoT has paved the way for Smart Cities,  while AI-fueled big data has revolutionized progressive urbanization. However, initiatives to promote technology must be balanced by principles of sustainability and livability. As deep learning has advanced rapidly, creating increasingly sophisticated technologies has led to highly complex — and often opaque — models that can be difficult to interpret. It becomes increasingly difficult to establish trust and maintain transparency when decision-making systems are based on such opaque and complex structures. This article explores the urban promise of AI and presents a new framework infusion of AI into cityscapes. The new direction is socially oriented through the inclusion of elements such as values, urban metabolism, and governance. A systematic review of machine-learning applications in cybersecurity also discusses the importance of explainability for overcoming the challenges it entails. The importance of assuring the explainability, interpretability, and intelligibility of autonomous systems will also be part of this discussion, especially in the context of developing smart cities using AI-based technologies.