With increasing dependence on electronic systems and sensitive data, effectively robust defenses are essential. This paper presents a machine learning-driven approach that utilizes face recognition techniques to safeguard an area from unauthorized access. The proposed system identifies access attempts and tracks users through machine learning and neural network algorithms to distinguish between legitimate users and potential intruder. The system employs anomaly detection in conjunction with face recognition to enhance security. A convolution neural network (CNN) is employed to extract facial components, ensuring accurate recognition even under challenging conditions such as occlusion or low illumination. For the gate system, an alarm system is enabled to trigger if a suspicious access is detected in the database, for the illegal faces, the algorithms of machine learning classifiers are trained on labeled datasets. Here, we extend the application of unreported access detection through facial recognition. In this paper detailed explicitly the action to be taken when unauthorized access is detected, which allows a legitimate company to track the illegal action that violates the security of the organization. Detection, Image Processing, Deep Learning, Unauthorized, CNN algorithm.