Transforming Network Security: The Role of Integrated AI in Intrusion Detection Systems | ||
| Advanced Sciences and Technology Journal | ||
| Volume 3, Issue 1, June 2026, Pages 1-14 PDF (218.21 K) | ||
| Document Type: Review Article | ||
| DOI: 10.21608/astj.2025.346317.1035 | ||
| Authors | ||
| Soha Safwat Labib* 1; Renad Samy Abdelsamad2; Mohamed Tarek Mahmoud2; Marina Magdy Khalil2; Mohamed Atef Abdelfatah2 | ||
| 1Computer Science, Faculty of Computer and Information Systems, Egyptian Chinese University, Cairo, Egypt | ||
| 2Software Engineering, Faculty of Engineering and Technology, Egyptian Chinese University, Cairo, Egypt | ||
| Abstract | ||
| The concept of Intrusion Detection Systems becomes very important in light of highly sophisticated and pervasive cyber threats. AI had finally become an agent of change by improving IDS systems through the application of sophisticated models of machine learning and deep learning in real-time complex security threat detection, analysis, and response. Besides, hybrid architectures that involve the integration of CNN and LSTM further apply AI-driven systems to make intrusion detection effective due to the capability to analyse even complex network traffic patterns. Combining spatial feature extraction with temporal sequence analysis makes it possible to precisely identify both known and unknown threats with high precision. Complex modern network environments detected the presence of obfuscated attacks and zero-day, which conventional methods could not detect. The incorporation of AI in IDS significantly enhances the detection accuracy while reducing false positives and offering adaptability in dynamic network conditions. These make the systems practical to run in real time, since analysis can be done without complex pre-processing of raw data; hence, they are indispensable for modern cybersecurity strategies. However, many challenges remain yet to be addressed regarding network traffic variability management, class imbalance handling, and false alarms minimization in an operational environment. Only continuous learning with model updates in this dynamic cyber threat environment will keep performance sustained. The present research has demonstrated the transformative role of AI in intrusion detection by underpinning its potential to revolutionize network security through proactive, adaptive, and resilient defences against an ever-evolving threat landscape. | ||
| Keywords | ||
| Malware; CNN; RNN; Machine Learning; Deep Learning | ||
|
Statistics Article View: 44 PDF Download: 5 |
||
