“It’s Up to Me Whether I Do—Or Don’t—Watch Deepfakes”: Deepfakes and Behavioral Intention

Abstract

The rapid spread of disinformation and misinformation created by deepfake poses threats and challenges to society. Most studies have analyzed deepfakes in terms of computer science, but there is now a pressing need to study them from the perspective of social science. This study aims to explore the factors that affect individuals’ intentions of watching or sharing deepfakes, the relationships between behaviors related to deepfakes and the impact of deepfakes to individuals or to the society. Accordingly, this study constructs a theoretical model and considers the impacts of deepfake content and quality on individuals’ behavioral intentions of deepfakes. This study adopts SPSS 26.0 to analyze the data set that involves 582 participants (Male = 234, Female = 348) in total. Results showed that attitude toward deepfakes, subjective norm, and perceived behavioral control positively influenced the individual’s behavioral intention regarding deepfakes, but perceived behavioral control did not have significant effects on behavioral tendencies on deepfakes when tested with other factors. Meanwhile, the quality of deepfake content was related to the individuals’ behavioral intentions of whether they were willing to watch or to share such videos, and there were gender, age, and education differences in individuals’ behavioral tendencies on deepfakes. Furthermore, the consequences of deepfakes were closely related to individuals’ sharing behaviors. The findings contribute to extended the TPB model, expanding the understanding of individuals’ differences in behavioral intentions of deepfakes. The findings also have implications for guiding the development of educational and training strategies to combat the negative effects of deepfakes.