Cloud computing systems are facing growing security concerns due to their distributed nature and continuously changing access behaviors. Conventional Identity and Access Management (IAM) approaches often fail to provide real-time responsiveness and lack awareness of contextual factors. To overcome these limitations, this study proposes a multi-layered security framework powered by Artificial Intelligence (AI). The model incorporates location-based dynamic code generation, user authentication, and machine learning–driven anomaly detection to strengthen security mechanisms.

The proposed system enhances authentication by integrating user credentials with geolocation information and time-dependent hexadecimal codes, making unauthorized access significantly more difficult. Furthermore, a trained machine learning model continuously monitors user behavior to identify and prevent suspicious activities. A practical scenario is included to demonstrate the effectiveness of the approach.

This review also examines the role of AI and Machine Learning (ML) in transforming cloud security, particularly in Identity and Access Management. It highlights the shortcomings of traditional access control systems in terms of scalability and real-time adaptability, and presents an AI-based framework capable of intelligent authentication, adaptive threat detection, and predictive decision-making. Based on the analysis of 34 research papers published between 2015 and 2025, including 29 journal articles, 1 article, and 4 conference papers, the study emphasizes that integrating AI with cloud computing is essential for developing secure, efficient, and context-aware access control systems. The proposed framework not only enhances security and reduces unauthorized access but also improves user experience and ensures better compliance with modern security standards.