<p>Services such as medical and disaster recovery are facilitated by cloud computing, along with corporate and new uses. Although public as well as private cloud operations are combined in hybrid cloud environments, security is a problem of this. Adequate safety standards cannot be achieved with the security procedures previously in existence. Thus, when maintaining, restoring, and accessing information in the cloud, this study offers a novel technique adaptive access control-driven optimal de-duplication with random forest (AAC-ODD + RF) for reducing possible risks to secure hybrid cloud networks. The advanced encryption standard (AES) technique and the MD5 hash function are the fundamental components of the optimal de-duplication (ODD) procedure developed in the suggested methodology, which ensures secure recovery and storage of information with no redundancy. The suggested security architecture incorporates the adaptive access control (AAC) approach to limit user data accessibility. Subsequently, the cloud users are classified using the random forest (RF) classifier. The suggested methodology was experimented and its security value was additionally assessed.</p>