In this presentation, we present MBBS, a tetra-model behavioral biometric-based authentication scheme designed specifically for smartphones. MBBS utilizes four distinct modalities to authenticate users: touchscreen swiping patterns, taps on “text-independent” 8-digit numbers, name writing on the touchscreen, and micro-movements of the hand during the entry process. To enhance overall accuracy and security, MBBS incorporates a state-of-the-art Generative Adversarial Network (GAN) powered data augmentation architecture. This innovative approach allows us to demonstrate the effectiveness of MBBS using both real user samples and augmented samples, consisting of a combination of “real” and “GAN-generated” data, on an actual Android device. One of the key advantages of MBBS is its high usability, as it eliminates the need for users to remember any secret information. Instead, it leverages users’ familiarity with natural processes, thereby increasing accuracy in real-time by employing GAN technology, all without requiring a large sample size from users. We will also present preliminary results from our performance, security, and usability analysis, which showcase a positive opinion regarding the effectiveness of our developed authentication mechanism.