Intrusion detection is a critical component within Internet of Things (IoT) environments, having role to safeguard interconnected devices, networks, and information from unauthorized access and malicious activities. As IoT continues to expand rapidly across domains such as smart homes, autonomous vehicles, industry, and healthcare, keeping these devices secure has become essential because of their resource constraints and different architectural designs. This research explores intrusion detection problem in IoT networks using the categorical boosting (CatBoost) classification model. A fresh variant of the recent chimp optimization algorithm is proposed to adjust the hyperparameters of the CatBoost classification model, enhancing its performance for intrusion detection within IoT networks. A thorough comparative experiment was carried out, comparing the introduced method with collection of other powerful optimization techniques within the same framework. The experimental results demonstrated the supreme performance of the introduced method, highlighting its significant potential in this specific application domain.