Survey on Artificial Intelligence for Cyber Threat Detection and Response in Cloud Environments

Abstract

The rapid expansion of cloud computing infrastructures has fundamentally transformed how organizations manage and deploy digital services, simultaneously introducing a complex and evolving attack surface that traditional security mechanisms fail to adequately address. This survey examines the convergence of artificial intelligence (AI) and autonomous cybersecurity with a focus on cloud environments. We systematically review thirteen recent papers spanning five core research themes: AI-driven threat detection and classification, explainable AI (XAI) for cybersecurity transparency, autonomous response and mitigation strategies, real-time cyber threat attribution, and AI-enhanced education for cybersecurity workforce development. Our analysis highlights the state-of-the-art techniques including Graph Neural Networks (GNNs), transformerbased attention mechanisms, Federated Deep Learning (FDL), reinforcement learning, and multi-modal data fusion, all applied to the challenge of building self-healing, autonomous cloud defense systems. We further discuss persistent challenges such as dataset quality, model interpretability, adversarial robustness, and the gap between academic research and real-world deployment. This survey provides a structured synthesis of the current state of the art, identifying key research directions for the next generation of intelligent, autonomous cloud security systems