Deepfakes: příležitost, nebo hrozba?

deepfakes

The term "deepfake" is a combination of the words "deep learning" and "fake" and represents a subset of synthetic media (voice, video, images, or a combination). These are created by deep neural networks and depict events that never happened. In addition to their potential positive uses, e.g. in the film industry and education, deepfakes have gained considerable attention mainly due to their potential negative effects, where they can be used to defame individuals, spread fake news, but also to attack biometric authentication systems or create phishing and scam sites. In this chapter, we introduce the different types of deepfakes, their creation and detection options. In explaining their application, we will also focus on the implications in the area of IT security.