The Internet of Things (IoT) and edge devices have seen widespread adoption across the world. Integrated and always-online devices, as well as those reliant on cloud support, have become a part of everyday life. However, due to their limited computational power and low cost, the security of IoT devices is often overlooked, with features being given much higher priority. As IoT devices can access the internet, they are susceptible to various types of cyberattacks. Moreover, their presence in homes and offices makes them valuable targets for malicious actors. This work proposes an artificial intelligence (AI)-based approach for detecting and identifying attacks within IoT networks. Given that the performance of classification models is highly dependent on proper hyperparameter selection, a modified metaheuristic algorithm based on the bat algorithm (BA) is introduced to optimize model performance. Simulations conducted on real-world data demonstrate promising results.