Recently cloud computing has become a fast growing area and helps to store very huge volume of data. In such a scenario, the computing resources including processor and memory are offered flexibly to the users as services on the internet. The users can get access to any of the resources available on the cloud based on their demand and they have to make payments for such services. In order to provide these features, access control techniques are essential since they satisfy the security requirements for handling the required payments effectively. In this paper, a new security mechanism for cloud computing with a focus on access control is proposed. In the past, different access control techniques including the traditional role-based access control (RBAC) model were used to implement the security features in databases. In this work, the traditional RBAC is modified to suit the security cloud requirements of databases by providing temporal extensions and intelligent agents. From the experiments carried out in this work, it has been observed that the proposed Intelligent Temporal RBAC is well suited to secure the cloud databases effectively. In this model, users or applications are clearly separated according to their job functions for database access with respect to the variation in time and situation. © 2014 IEEE.