Abstract
The development of artificial intelligence-based communication technologies carries significant risks in the relationship between human and machine, in means of information manipulation. Deepfake creates a world where the technology in question is far from being “artificial”, as the name of the technology it has created and perceived as real day by day. Deepfake, which is defined as artificial intelligence-based visual disinformation, has brought the problem of reliability inherent to information. For this reason, it is deemed to evaluate the literature that discusses the risks of disinformation content sharing on information manipulation with a deepfake perspective, which can be considered as a form of visual disinformation. Deepfake has made it possible to mass-produce synthetic videos that are very similar to real videos. It has become possible to reach deepfake videos containing political discourse in countries other than Turkey. The starting point of this study is to present a comprehensive contribution in the context of the literature concerning the disinformation deepfake in Turkey. From this point of view, the technology that forms the basis of the deepfake creation process as a form of visual disinformation has been discussed, and the methods that enable deepfake technology to manipulate the audio-visual media are presented. After compiling the information that will enable us to understand how deepfake videos work, the dilemmas of deepfake-based manipulative contents in the special context of political communication are discussed.