CheapVersus Deep Manipulation:The Effects of CheapfakesVersus Deepfakes in a Political Setting

Abstract

Visual disinformation has been regarded as convincing because it strongly resembles reality. Yet, we lack a clear understanding of the effects of different forms of audiovisual disinformation—cheapfakes versus deepfakes. To advance the disinformation literature, this paper reports on the findings of two experiments in which participants were exposed to political cheapfakes and deepfakes, respectively. Our main findings indicate that audiovisual disinformation is not perceived as more credible or believable than the same disinformation in textual format. Importantly, deepfakes are perceived as less credible than cheapfakes with a similar de-legitimizing anti-immigration narrative. Although more research is needed, our findings suggest that less sophisticated modes of deception can be at least as credible as more sophisticated forms of artificial intelligence-driven audiovisual fabrication.