Enhancing Transparency in Cloud-Based Access Control: Analyzing RBAC, ABAC, DLBAC, and XAI Techniques

Cloud computing has emerged as a dominant paradigm for delivering computing resources on demand, enabling organizations to scale operations rapidly while reducing infrastructure and maintenance costs. Its widespread adoption across sectors such as healthcare, finance, education, and public administration has transformed how digital services are deployed and managed. At the same time, the shared, distributed, and multi-tenant nature of cloud platforms introduces complex security challenges, particularly in access control, where unauthorized or inappropriate access can lead to serious operational and compliance risks. Traditional authorization mechanisms, including Role-Based Access Control (RBAC) and Attribute-Based Access Control (ABAC), provide structured, policy-driven approaches for managing permissions. Although effective in relatively stable environments, these models often lack the adaptability and contextual intelligence required for dynamic cloud ecosystems characterized by changing user behaviour, diverse devices, and evolving threat patterns. Recent advances in Artificial Intelligence (AI), especially Deep Learning–Based Access Control (DLBAC), have improved decision accuracy and enhanced the detection of anomalous access behaviour. However, such AI-driven systems typically operate as black boxes, offering limited insight into how access decisions are made. This opacity raises concerns related to transparency, trust, accountability, auditability, and regulatory compliance. This paper presents a systematic analysis of RBAC, ABAC, and DLBAC models, alongside Explainable Artificial Intelligence (XAI) techniques, including SHAP and LIME. By integrating explainability into access control decision-making, the study identifies pathways toward more transparent, responsible, and trustworthy cloud security architectures that align with modern sustainability and compliance requirements.