Predominance of cheapfakes in social networks Technical complexity and textual functions of disinformation checked in Argentina during 2020

ABSTRACT

Research on disinformation concentrates a growing interest in the design of deepfakes and the possible political, economic and social implications of their viral dissemination on social networks. However, as much or more worrying are the cheapfakes the hoaxes created by the users themselves through the native features of their mobile devices, the rude manipulation of pre-existing files or the simple addition of text that alters the original meaning of the shared messages. The objective of this research is to analyze the degree of technological complexity of the disinformation that circulated in Argentina during 2020 throughout social networks, as well as the role that the text played in its creation. For this, a content analysis is applied to a corpus made up of the 201 denials referring to the disinformative material that was released during that year by the fact-checking platform Chequeado, the only one accredited by the International Fact-Checking Network in the country. With the results we propose a curve of technical complexity of disinformation that shows the absolute predominance of hoaxes associated with very low, low or medium digital skills. This proves that cheapfakes are much more popular on social media than those desinformative content created professionally or through artificial intelligence. In addition, it is found that falsehoods are essentially created through captions or the insertion of text in the images, whose main function is to disinform about events or specific facts through the dissemination of wrong data or the omission of other true ones.