The vogue and advanced functionality of mobile devices to create them striking targets for malevolent and invasive applications (apps). They need security measures are in the own place for mobile system. The user to make decisions that crash the security of a device, where these system often to fail. Say for example, Android depends on users to identify the permissions that an app is requesting and to base the fitting decisions on the list of permissions. In the proposed model, the authors creating a monitoring neural learning for a software verifications. This model to tracked, if there any misbehaved application accessed by user it will be blocked it.