Security has always been a major problem with cloud computing. The most important and flimsy data that attackers point victims toward are information sources. Each cloud client’s security and protection are compromised in the unlikely event that information is lost. A cloud network is monitored remotely, yet there is still a risk of an intruder from within. Interior attackers obtain entry to a framework by haggling over a feeble client hub. They launch attacks pretending to be confidential client information and are connected to the internal cloud network. AI techniques are typically applied to cloud security problems. According to the current AI-based security techniques, a hub is described as being troubled when it becomes dependent on sporadic behavioral data. These frameworks do not distinguish between a resentful hub and an out-of-control hub. This work suggests an improvised model that learns a client’s behavior, automatically gets ready, and archives social media data in order to solve this problem. Without a doubt, the model can classify the client behavior as normal or unusual. Using the determined trust factor, the suggested approach finds an oddity hub and uses it to identify whether an acting mischievously hub is a wrecked hub, another client hub, or an undermined hub. The suggested model reduces the false alarm in the cloud organization and accurately identifies the attack.