Detection of Stealthy Encryption in Ransomware Using AI-Driven Anomaly Detection Models

Abstract Ransomware continues to pose a significant threat to cybersecurity, with increasingly sophisticated techniques allowing malicious actors to evade traditional detection mechanisms and inflict substantial damage on both individual and organizational levels. The introduction of an AI-driven detection model that integrates anomaly detection with supervised learning offers a novel approach to identifying ransomware activities, particularly those utilizing stealthy encryption techniques that are designed to avoid detection. Through comprehensive evaluation, the proposed model has demonstrated superior performance compared to existing methods, achieving higher detection accuracy, reduced false positives, and enhanced resilience against adversarial evasion. The model's scalability and efficiency across diverse operational environments further demonstrate its practical applicability, making it a viable solution for real-time ransomware detection in both high-performance and resource-constrained settings. The research contributes to the ongoing efforts to fortify cybersecurity defenses by offering a robust, adaptable, and scalable framework capable of addressing the evolving nature of ransomware threats.