Increasingly, the need for reliable verification has brought the study of multi- modal biometric systems integrating fingerprints, faces, irises, and veins. These systems provide more security, accuracy, and resistance to spoofing compared to single-mode procedures. Moreover, the continued efforts in advancing cryptographic methods such as AES encryption, homomorphic encryption, and the use of blockchains intensify the security of biometric templates from cyber threats. Unfortunately, such hopeful advancements still leave standardizing evaluation metrics, optimizing computational efficiency, and addressing privacy-threatening challenges. This study examines the new advancements in multi-biometric authentication focusing on cryptographic components, Advanced Encryption Standard (AES), Elliptic Curve Cryptography (ECC), and Secure Hash Algorithm 256 (SHA-256) watermarking, multifactor authentication, and combinatorial fingerprinting. After significant research, the achieved results are biometric recognition accuracies of 92-96.67% and mobile app error rates of below 1%. The various ways biometrics can increase accuracy and then reduce the incidences of false acceptance and rejection aid the enhancement of anti-fraud protection in a multi-biometric system. Nevertheless, in this scenario, numerous challenges remain, such as for uniform evaluation, computational cost, and privacy issues present hurdles to the multi-biometric paradigm. This review discusses literature, states the existing gaps, and recommends methodologies such as blockchain, quantum cryptography, and real-time authentication for future research on mainstream biometric systems. It also seeks to advance biometric systems used in e-learning, finance, and national security by evaluating their security parameters against new technologies.