<p>With exponential growth, the adoption of digital wallets, or e-wallets, is increasing as a result of efficiency and convenience. Thus, this large utilization of e-wallets will need secure authentication of users, to preclude unauthorized access and fraud. In this paper, the current authentication methods for e-wallets, including knowledge-based authentication (KBA), possession-based authentication (PBA), biometric-based authentication (BBA) and multi-factor authentication (MFA), are discussed. We also examine strengths and weaknesses of these approaches in the most popular e-wallet systems globally. In addition, a discussion of possible threats to e-wallets, including a password or PIN theft, cracking passwords, stealing biometric data, theft of devices, man-in-the-middle attacks, malware attacks, and session hijacking is considered. We present an automated tool VoicePass that combines methods of KBA, PBA, and BBA with voice recognition to make e-wallet transactions safer. VoicePass authentication process is based on the entry of a password as the first factor of authentication, receiving of one time password either through SMS or dedicated authentication app, speaking and pronouncing the OTP to the e-wallet system and confirming the user’s identity by using the voice recognition technology and grant We use actual dataset (i.e., AudioMNIST) containing 60 loudspeakers in our simulated user behaviours as well as sound patterns. We analyze performance of VoicePass and demonstrate that it achieves an authentication accuracy rate of 99.64% suggesting that VoicePass can identify and authenticate user authentication. Our tool also yields precision and recall values of 0.9965, and 0.9964, indicating the success of the tool in identifying authentic authentication efforts and reducing errors. The F1 score value (0.9953) is a harmonic measure of precision and recall and points to the balance of the two.</p>