The rise of the Internet of Things has brought about various advantages, such as providing us with more efficient and effortless activities. Unfortunately, the lack of security solutions has also led to the development of new threats. One of these is the exploitation of vulnerabilities in the networks of IoT devices. In order to effectively address the security threats that can arise in the networks of IoT devices, there needs to be an effective intrusion detection system (IDS). In the field of security, the use of artificial intelligence (AI) powered IDS has shown promising promise. Through deep learning and machine learning techniques, these systems can learn and adapt quickly to new threats. This paper presents an evaluation of the performance of an AI-based security system on a large dataset. The research begins with a literature review of the previous studies related to the security of IoT devices and intrusion detection. We then develop a methodology that includes the data collected for evaluation and training, an AI model architecture for intrusion detection, and the evaluation metrics. The paper presents the results of the study and discusses the performance of the AI-based IDS compared to the existing solutions for addressing security threats in Internet of Things networks. It also explores the potential of this technology for future research. The findings of this study contribute to the growing body of research on the security of IoT networks and intrusion detection. It shows that an AI-based IDS can perform better than the existing solutions in identifying and mitigating threats. The study's findings show the potential of deep learning and machine learning techniques to enhance the security of IoT networks. It also highlights the scope of this technology's application in other security domains.