In the increasing reliance on the cloud computing era, securing digital assets against sophisticated cyber threats has become a critical concern for organizations globally. Traditional security mechanisms, which often rely on static and pre-defined access control policies, must be revised to address these threats' dynamic and evolving nature. This study investigates the application of Artificial Intelligence (AI) and Machine Learning (ML) in enhancing cloud security through the development of advanced Risk-Based Access Management (RBAM) systems. The primary objective is to evaluate how AI and ML can improve dynamic access control, threat prediction, and mitigation strategies within cloud environments. The research adopts a mixed-methods approach, combining quantitative analysis of RBAM system performance with qualitative insights from cybersecurity experts. AI/ML models were developed using extensive historical access log datasets and integrated into a cloud-based RBAM prototype. The system's performance was assessed based on its accuracy in threat detection, reduction in false positives, and effectiveness in dynamically adjusting access controls. Results indicate that the AI-enhanced RBAM system significantly outperforms traditional methods, achieving a 30% reduction in false positives and a 25% decrease in unauthorized access incidents. Additionally, AI-driven threat prediction models demonstrated high accuracy, enabling preemptive actions to mitigate potential security breaches. These findings highlight the transformative potential of AI and ML in cloud security, providing a more adaptive and proactive defense against emerging threats. The study concludes with recommendations for refining AI/ML models and exploring their application in other areas of cloud security, emphasizing the need for continued innovation to safeguard the increasingly complex digital landscapes that organizations operate within today.