Deepfake Attacks on Biometric Recognition: Evaluation of Resistance to Injection Attacks

Deepfake technologies, despite advancements in facial and voice recognition, pose significant threats by replicating voices or faces, compromising identity verification processes even during live interactions. This creates risks of impersonation and unauthorized access to sensitive data. Addressing these threats is crucial, especially given the ease of creating deepfakes with free open-source apps, which have advanced to the point where distinguishing between real and fake is increasingly challenging. Our developed system enhances biometric security by leveraging existing hardware, specifically screen illumination, to scrutinize light reflections from the user’s face. This method seamlessly integrates into current applications, adding a necessary layer of security to differentiate genuine individuals from deepfakes, thereby fortifying the overall biometric security framework without requiring additional algorithms or hardware.