Abstract

This study surveyed 628 Vietnamese youth to analyze the relationship between perceptions of deepfake technology and behavioral responses on social media. A Likert scale questionnaire was designed to investigate and measure concepts related to situational discernment, perceived persuasiveness/influence, perceived risk, trust in platforms and technology, subjective norms, and perceived behavioral control. After reliability testing, exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) confirmed acceptable measurement values for the refined scale model (with two indicators for each concept). Descriptive analysis revealed a high and consistent perception of the influence of deepfake videos, while confidence in discernment was at a moderate level with significant variability. Regarding behaviors in the past six months, sharing and warning were reported more frequently than reporting. Logistic regression analysis showed that perceived risk predicted a decrease in the likelihood of denial/warning comments, while trust in platforms/technology and perceived behavioral control had only small positive correlations with warnings. The results suggest that conceptual awareness has matured, but the transformation into systematic defensive behavior remains limited due to a lack of discernment confidence and responsive behavior on platforms.