Political deepfakes have been regarded as potentially harmful for democracy. However, to date, we lack a clear understanding of how people respond to deepfakes, especially when they differ in plausibility and the extremity of the deceptive message. Against this backdrop, this study relies on a qualitative and quantitative analysis of open-ended questions (N = 601) asked after exposure to 3 different political deepfakes. These deepfakes either contained a plausible message or implausible messages with counterattitudinal statements. Distrust in the plausible deepfake was mostly attributed to the insincerity of the depicted politician. The implausible deepfakes, in contrast, were also doubted because of perceived issues with the authenticity of the footage. As a main contribution to the (AI-powered) disinformation literature, we show that the credibility of disinformation is understood in different ways depending on the context of deception, ranging from the plausibility of content to the authenticity of the presentation of synthetic media.
