Exploring the Impact of Synthetic Political Video on Citizens’ Trust in News

• In Event: Visual Frontiers in Digital Politics

Abstract

Artificial Intelligence (AI) now enables the mass creation of what have become known as “deepfakes”: synthetic videos so realistic that the human eye cannot reliably distinguish them from real ones. To shed light on the potential perils of political deepfakes we administered novel experimental treatments to a large online sample representative of the UK population. Using different cuts of an educational video produced by news organization BuzzFeed in 2018 depicting a synthetic Barack Obama calling President Donald Trump a “dipshit,” we compare subjects’ evaluations of deepfakes that do and do not reveal themselves as fake. We find that people are more likely to be confused rather than misled by deceptive deepfakes. We develop an account of why this is the case, based on our interpretation of what makes political deepfakes unique forms of disinformation. We show how confusion, in turn, reduces levels of trust in news on social media, creating problematic implications for the quality of online public discourse. Our findings suggest that deepfakes may erode citizens’ certainty and trust, further intensifying recent challenges to online civic culture in democratic societies.

Authors

• Cristian Vaccari, Loughborough University

• Andrew Chadwick, Loughborough University