Integrating Deepfake Detection into Cybersecurity Curriculum

Deep Learning (DL) has been widely adopted in the domain of cybersecurity to address a variety of security and privacy concerns. Moreover, in recent years attackers are also increasingly adopting deep learning to either develop new sophisticated DL-based security attacks, such as Deepfakes. Recently Deepfake technology is used to spread misinformation on social networking. Currently, the most popular algorithm for deepfake image generation is GANs. The goal of this paper is to adopt DL-based smart detection techniques to defend against smart misinformation. We develop a set of hands-on labs to integrate them in our cybersecurity curriculum so that our students, future cybersecurity professionals, can be educated to use detect software and identify Deepfakes. Finally, we will investigate the fundamental capabilities, challenges, and limitations of deep learning for detecting smart attacks.