International Research Journal of Education and Technology

       With the rise of sophisticated cyber threats, traditional Intrusion Detection Systems (IDS) struggle to detect evolving attacks, leading to security vulnerabilities. This study proposes an AI-powered IDS that leverages machine learning (ML) and deep learning (DL) to enhance network security. The system utilizes Random Forest, SVM, LSTM, and a CNN-LSTM hybrid model to analyze network traffic and detect anomalies. Trained on datasets like NSL-KDD, CICIDS2017, and UNSW-NB15, the proposed model improves accuracy and reduces false positives compared to conventional IDS. Despite challenges such as computational complexity and data imbalance, AI-driven IDS offers a promising solution for real-time secure network communication.