Motivations for (not) sharing deepfakes on social media

Abstract

Deepfakes–AI-generated media that manipulate images, videos, or audio to create realistic but false content–are increasingly viral, making it crucial to understand their sharing dynamics and help society navigate the challenges they pose. This research applies Uses and Gratifications theory to examine individual motivations for sharing or not sharing deepfakes. A total of 1,035 US participants (42% female; Age: M = 42.8, SD = 14.1, Range = 18–93) watched one of ten deepfakes from five different genres and completed a questionnaire on Prolific. Results show that most participants would not share the deepfakes due to a lack of personal relevance. However, those who intended to share them had hedonistic motivations, primarily enjoyment and the desire to share entertainment with others. The genre of the deepfakes influenced motivations to share, with political deepfakes being a particularly specific case, where gender also had a greater impact. Women were more likely to avoid sharing inaccurate information, while men were more focused on protecting their social image. This research emphasizes the significance of perceived entertainment value in driving deepfake virality and recommends studying both motivations for sharing and not sharing content.