Leveraging Deep Learning for Access Control and Data Security in Cloud Environments

This is the authors' abstract. We don't add key points for this paper.

Cloud computing emerged as a promising model for delivering services over the internet, offering users a pay-as-you-use structure that grants access to infrastructures (Infrastructure as a service), platforms for running various programs (Platform as a Service) and software (Software as a Service) at minimal cost. Surprisingly, as cloud environments evolve and expand, access control has become a critical concern to ensure that the data stored in these systems maintains confidentiality, integrity, and availability. The increasing complexity of cloud architecture highlights the urgent need for a robust access control model, as the consequences of inadequate protection could be disastrous. This article provides a novel approach to addressing the vulnerabilities in cloud security, emphasizing the need for enhanced access control measures. It investigates how deep learning can provide a defense-in-depth strategy to detect and prevent unauthorized access and data breaches by analyzing current gaps in cloud security research. Through a systematic analysis, this study aims to bridge these gaps, offering insights into how deep learning models can strengthen cloud environments against potential threats. By combining deep learning with traditional methods, the model dynamically optimizes access control decisions and strengthens data confidentiality. Results show that the proposed outperforms current works with accuracy of 94.39% that demonstrates considerable gains in data security, resilience, and overall performance that effectively addresses anomalies in cloud environments.