Cyber-Physical Systems seamlessly integrate computation, communication, and control with physical processes across key sectors such as healthcare, transportation, and energy. Applications in these domains—ranging from surgical robots and patient monitoring to self-driving cars, drone traffic control and smart power grids—have revolutionized modern infrastructure. However, the growing complexity and interconnectivity of CPS make them highly vulnerable to sophisticated cyber threats. Artificial Intelligence (AI)-driven security solutions have emerged as a promising approach to safeguarding these critical systems. Yet, most AI security solutions operate as black-box models, limiting transparency, trust, and widespread adoption in safety-critical applications. This chapter provides a comprehensive analysis of AI-driven security frameworks for CPS, emphasizing the need for explainable AI (XAI) techniques to enhance interpretability, reliability, and user confidence. By bridging this gap, researchers, practitioners, and policymakers can design and implement more accountable and trustworthy CPS security solutions. Finally, this chapter explores future research directions, advocating for the integration of XAI methodologies to ensure robust, interpretable, and ethical AI applications in CPS security.