Who Are the Targets of Deepfakes? Evidence From Flagged Videos on Tiktok and Youtube

Abstract

This study draws on a dataset of 1413 (openly) doctored videos from YouTube and TikTok, coded by the race, gender, profession and nationality of the individuals featured in the clips. The findings suggest that, by stark contrast to pornographic sites, deepfakes shared on mainstream platforms overwhelmingly feature individuals who are white, male, and—when in politics—authoritarian. Based on a qualitative analysis of the videos, it appears that their chief cultural function is not satire but what may rather be interpreted as a carnivalesque profanation of the faces of power. These findings challenge previous studies on the demographics of deepfake targets and broaden the scope of current literature by shedding light on how deepfakes are used beyond deception and porn.

Keywords: Deepfakes, Generative AI, TikTok, YouTube, Social Media, Satire, Disinformation