AI-Based Secure Vault System Using AES-256 and Biometric Authentication in IoT

Abstract

In this paper, the design of the AI-based Secure File Vault incorporating multimodal biometric identification alongside AES-256 encryption is presented. Conventional passwords and other forms of authentication suffer from vulnerabilities to phishing attacks, making it necessary for the system to use facial identification and voice recognition technology for user authentication prior to access. AES-256 encryption ensures security of the data. The design incorporates session decryption, auto-encryption, and application-level locking to ensure further security. Results show that the face recognition and voice recognition rate is about 98% to 99% and 95% to 97% respectively, with an authentication speed of less than a second.