A new taxonomy for image use in the intentional shaping of the digital narrative: deep fakes and artificial intelligence

Abstract

Any confrontation seeks the (partial) imposition of one party´s will on an adversary. In recent years, digital technologies and data science have combined to create new ways of controlling the narrative and carrying out powerful information or disinformation campaigns that are part of a new type of warfare with digital characteristics. In particular, based on a literature review and the direct work of authors in different expert groups, this paper studies the use of either static or moving images (real, altered, or wholly artificially generated) as one of the most efficient means of altering perceptions and thereby narratives in the so-called cognitive domain. To this aim, this article collects in an orderly and exhaustive way the most recent military and intelligence doctrine related to such "influence operations," from a dual technological-sociological perspective. Based on this, the paper determines the differences between traditional disinformation techniques and those incorporating digital and artificial intelligence technologies in the form of images and video. The paper proposes a new double-entry taxonomy that can display the degree of image faking and the objective of disinformation. This helps to identify and prioritize the most relevant cases and thus adopt the most appropriate countermeasures. These are also examined in detail, leading to the conclusion that only a combination of transparency, consumer education, technology, and regulation can counteract the increasing use of images and video with false content.