PRIVILEGE SHADOWS: INTELLIGENT DETECTION OF ESCALATION THREATS IN CLOUD INFRASTRUCTURE

Privilege escalation attacks in cloud environments pose serious security threats, compromising the integrity and confidentiality of sensitive data. These attacks occur when unauthorized users gain elevated access rights, leading to significant organizational risks. Traditionally, detection relied on manual monitoring and rule-based systems with fixed thresholds and basic anomaly detection, which were often slow, inefficient, and dependent on human oversight. As cloud infrastructures grow more complex and attackers become more sophisticated, these traditional approaches have proven inadequate. To address these challenges, the use of machine learning has emerged as a powerful alternative, enabling automated detection of abnormal behavior with improved accuracy and scalability. This research focuses on leveraging machine learning models such as Random Forest, AdaBoost, XGBoost, LightGBM, and CatBoost to detect and mitigate privilege escalation attempts. These models are trained on historical attack data, allowing them to identify complex patterns that traditional systems often miss. Unlike older methods that suffer from high false positives and delayed responses, the proposed system uses advanced classification algorithms to provide real-time, precise detection of malicious activities. By continuously learning and adapting to new threats, the system offers a robust, scalable, and proactive security solution. The research highlights the effectiveness of machine learning in enhancing cloud security, particularly in mitigating privilege escalation attacks, and demonstrates significant improvements over conventional techniques.