Location-based AES Encryption and XGBoost Classification in a Hybrid Access Control System

In order to improve cybersecurity, this project offers a strong multi-factor access control system that combines machine learning-based decision-making, geolocation-based validation, and AES encryption. Access is only provided to users who are physically present in designated office zones, as confirmed by IP-based geolocation. To ensure confidentiality, messages are encrypted in CBC mode using AES. We create behavioral access records and train an XGBoost classifier to predict access permissions based on user role, department, login frequency, and other factors in order to further impose intelligent access control. The system handles security issues such location spoofing, brute-force attacks, and illegal access while achieving high accuracy and interpretability. For a more intelligent, context-aware access control system, our solution shows how well classical encryption works when combined with real-time data analytics and artificial intelligence.