XAI-Driven Approaches for Ensuring Security and Data Protection in IoT
Amita Sharma, Navneet Sharma, Anubha Jain · 2024
Explainable artificial intelligence (XAI) emerges as a promising solution to address the complex security challenges posed by the Internet of Things (IoT). Unlike traditional black box AI, XAI provides transparency and interpretability, granting us a clear understanding of how AI models arrive at their decisions within IoT security applications. This newfound transparency allows us to identify potential vulnerabilities and threats more effectively, empowering us to implement proactive measures that safeguard IoT systems and protect data privacy. By embracing XAI, we can transcend the limitations of black box AI, ensuring a more secure and trustworthy IoT landscape for the future. This chapter explores the convergence of IoT, AI, and XAI, focusing on security challenges in IoT systems. It highlights the applications of AI in IoT, while emphasizing the significance of XAI in providing transparency. The chapter examines security issues addressed by XAI, supported by case studies in smart irrigation, drug design, and many more. We will also discuss explainable multi-agent systems developed for IoT-based problems. Ethical considerations and future prospects are also discussed.