Machine and deep learning-based intrusion detection systems (IDSs) have demonstrated superior learning performance for the Internet of Things (IoT) environment. However, the issue with some state-of-the-art models is their lack of explainability, transparency, and trustworthiness. Explainable artificial intelligence (XAI) has emerged as a solution to overcome this issue. It aims to make black-box models, such as deep learning models, more intelligible to end-users. This paper explores recent advances in XAI-driven intrusion detection systems tailored for IoT environments. In addition, it addresses the unique challenges that IoT systems pose, such as scalability and the trade-off between accuracy and explainability. It highlights promising future research directions, including hybrid models. By surveying the current landscape of XAI in IoT intrusion detection, this work aims to guide researchers and practitioners in developing more robust, transparent, and trustworthy security IDS solutions for IoT ecosystems.