Cybersecurity in AI-Enabled IoT Network: Threat Detection and Mitigation Using Deep Learning

Abstract

The rapid evolution of Artificial Intelligence (AI) and the Internet of Things (IoT) has reshaped the contemporary digital ecosystem, enabling the introduction of smart environments across most disciplines. However, this integration has raised serious cybersecurity concerns due to massive connectivity, diverse architectures, and the real-time flow of information. The article proposes a hybrid CNN-based LSTM system for threat detection and mitigation in AI-operated IoT networks. The proposed system is tested using the CICIDS2017 dataset, a popular intrusion detection benchmark. The model performs feature extraction, anomaly detection, and automatic mitigation to enhance cybersecurity resilience. Experimental results demonstrate that the proposed CNN-LSTM model achieves 96.8% accuracy, outperforming traditional machine learning models such as Decision Tree, SVM, and KNN. The model also achieves lower false-positive rates and improved detection time, making it useful for real-time security applications in the Iot. The results show that deep learning methods are highly efficient in improving Iot security and achieving maximum protection against emerging threats.