Cloud computing faces numerous challenges in many areas including security and privacy issues. In this work, a developed approach is suggested to tackle three security and privacy issues: network intrusion detection (NID), privacy, and internal attacks. A decision tree (J48) has been used to generate a set of rules based on the CICIDS2017 dataset to solve the NID problem. The accuracy of the generated rules approaches 99.8%. A set of policies are attached to the data file on the bases of a sticky policy to preserve privacy. A new approach is suggested based on blockchain to detect internal attacks in real-time, in which a set of trustees-chain are identified by the data owner. Any data modification conducted by a trusted member will be reported to all members of the trust group including the owner. The developed approach suggests adding a Privacy and Detecting Intrusions Service (PDIS) layer as part of the cloud computing main service model. PDIS includes the three suggested approaches above (NID, sticky policy, and trustees-chain). Finally, a web-based application is implemented to act as casework to validate PDIS and evaluate its reliability. Keywords: Cloud computing, Privacy, Machine learning, Internal Intrusion Detection, Network Intrusion Detection, Real Time Auditing