An AI-based Online E-Voting System is developed to provide a secure, transparent, and reliable digital voting mechanism by integrating face recognition techniques with Java and SQL-based processing. The system authenticates voters by capturing live facial images and comparing them with registered facial data using machine learning and computer vision methods to prevent impersonation and duplicate voting. It validates voter eligibility, enforces one-time voting through database constraints, and securely records votes to ensure data integrity and accuracy. Users interact with the system through a user-friendly interface where voter registration, authentication, and vote casting are performed seamlessly. The backend application processes voting requests, manages election data, and automates vote counting and result generation. By leveraging AI-driven facial authentication instead of traditional credential-based verification, the system enhances election security and minimizes manual intervention. The proposed framework improves the efficiency, trustworthiness, and scalability of online voting systems and supports fair and reliable elections in institutional and organizational environments.