Users? emotional and behavioral responses to deepfake videos of K-pop idols

Abstract

Deepfake is a process and outcome of artificial intelligence (AI) to create hyper-realistic manipulated media content. Despite its benefits, great attention is paid to deepfake pornographic content, which is a tremendous threat to media figures and the public. This study thus explains how general online users feel about and react to sexual deepfakes, focusing explicitly on K-pop idols. We conducted an online survey with individuals who have experience watching any deepfake content, and we analyzed the model using the PLS-SEM method by SmartPLS 3.0. The results showed that previous perceptions about sexual harassment and K-pop idols, experiences viewing pornographic content, and gender were crucial predictors of viewers' emotions toward K-pop idols' deepfake porn videos. Media exposure and age, however, did not have significant effects on users' emotional responses. The results also indicated the substantial positive effects of anger on the coping behaviors of problem solving and emotional support. Guilt, on the other hand, was not associated with either behavioral response measured. This study represents the first attempt to empirically investigate individuals' negative feelings toward and reactions to content generated by novel and maliciously exploited technology. It also highlights the necessity of collaboration among government, industry, and education to address deepfakes of a sexual nature.