ADAPTIVE MACHINE LEARNING FOR CONTINUOUS SECURITY TESTING IN CLOUD-NATIVE SOFTWARE SYSTEMS

Abstract

Cloud-native software systems have transformed modern application development through scalable architectures, microservices, containerization, and continuous delivery practices. However, the dynamic nature of cloud environments introduces complex security challenges, including evolving vulnerabilities, misconfigurations, dependency risks, and rapidly changing attack surfaces. Traditional security testing approaches often lack adaptability and struggle to provide continuous protection in highly dynamic cloud-native ecosystems. This paper proposes an Adaptive Machine Learning Framework for Continuous Security Testing in Cloud-Native Software Systems by integrating adaptive learning algorithms, DevSecOps practices, automated security testing, vulnerability intelligence, and cloud-native monitoring mechanisms. The proposed framework continuously analyzes application behavior, source code changes, infrastructure configurations, security events, and testing outcomes to detect emerging security risks. Adaptive machine learning models dynamically update their knowledge based on new threats and system changes. Experimental evaluation demonstrates improvements in vulnerability detection accuracy, threat adaptation capability, continuous monitoring efficiency, and cloudnative security resilience. The proposed framework provides an intelligent approach for maintaining secure cloud-native software environments. Keywords—Adaptive Machine Learning, CloudNative Security, Continuous Security Testing, DevSecOps, Vulnerability Detection, Automated Testing, Artificial Intelligence, Software Security.