A Photo and a Few Spoken Words Is All It Needs?! On the Challenges of Targeted Deepfake Attacks and their Detection

With the rise of artificial intelligence (AI) in recent years, AI has not only become more accessible to the general public but also presents new challenges. Images, videos, and audio can now be easily created or altered using deepfakes, making them a threat to human society. There is a high risk of being exploited for fraud and defamation, particularly with targeted deepfakes, a combination of social engineering and deepfake technology. It is therefore of great importance that such attacks can be detected. In this paper, we provide an overview of various techniques for creating such targeted deepfakes and discuss forensic and organizational measures to detect them as well as the common challenges associated with them.