<p>The rapid growth in the areas of Internet of Things (IoT) and Wireless Sensor Networks (WSNs) has made a huge impact on almost all the domain areas, specifically healthcare, smart factories, and transportation, as well as environmental monitoring. These massive networks connect millions of devices, but this large scale integration of diverse devices makes these networks complex, and their diverse architectures and limited computational resources make them highly vulnerable to a wide range of security attacks. AI and ML are extensively used as essential tools to safeguard these networks, as AI and ML have the capability of identifying diverse patterns, detecting unusual behavior, blocking attacks, and can adapt or adjust to new threats. Although AI and ML models provide strong predictive capability but many of the ML models work like Black Boxes; they make decisions but do not state how they have taken that decision and why a particular parameter is selected in the decision making. In a high risk environment, the black box nature or the lack of transparency of models becomes a major drawback because security decisions must be understandable, traceable, and trustworthy. These limitations of the existing models led to the rise of Explainable AI (XAI). Unlike previous surveys that focus on general XAI frameworks, this work examines explainability techniques specifically for security in IoT and WSN environments.</p> Graphical Abstract <p></p>