Artificial intelligence, disinformation and media literacy proposals around deepfakes

Abstract

The role of artificial intelligence and its place in the new disinformation strategies is perhaps one of the most difficult issues to focus on nowadays, since we are at the beginning of a process of definition and ways of exploration. In this paper, first of all, we analyze the different approaches that are being applied to the regulation of artificial intelligence and that may affect the different disinformation strategies that are being identified. Secondly, we study how artificial intelligence is being used to identify disinformation content. In this regard, from the point of view of verification processes, one of the main challenges is when identifying deepfakes (images and video, mainly) linked to news cycles. From this perspective, a typology of deepfakes is proposed and its main characteristics will be described according to the verifications carried out by the Spanish fact-checking organizations. Finally, a set of recommendations will be presented to work from a media literacy point of view with the identification of deepfakes.