Published Paper Details:

The rapid escalation of cyber threats, characterized by sophisticated zero-day exploits and polymorphic attacks, mandates a transition from traditional reactive security postures to proactive, AI-driven defense mechanisms. This paper presents a comprehensive systematic review of the state-of-the-art applications of Artificial Intelligence (AI), Machine Learning (ML), and Deep Learning (DL) in enhancing cybersecurity resilience. We analyze core defensive architectures, focusing on Network Intrusion Detection Systems (NIDS) and advanced malware classification, providing a quantitative comparative analysis of model performance across critical metrics such as Accuracy and Inference Duration. Our findings highlight the superior detection capability of Deep Learning models, specifically Transformer architectures, which achieve up to 98.4% accuracy in NIDS benchmarking. Furthermore, this review critically investigates the dual use challenge, examining emergent adversarial risks, including data poisoning, evasion attacks, and unique vulnerabilities introduced by Generative AI (GenAI), such as prompt injection and the phenomenon of feedback loop security degradation during automated code generation. Finally, we establish the mandate for Explainable AI (XAI) and robust governance frameworks as essential components for ensuring transparency, accountability, and trustworthy deployment of autonomous security systems.