Immunizing the Public Against AI-Generated Disinformation: Testing the Effects of Inoculation Mode and Issue Attitude on Inoculation Likelihood of Political Deepfakes

Abstract

Political deepfakes are considered detrimental to democracy by eroding public trust and distorting communication. Scholars have advocated for inoculation strategies to counter deepfakes, yet they have found that individuals’ partisan attitudes can undermine the effects of inoculation. Guided by inoculation theory and motivated reasoning theory, we conducted a 3 (Inoculation Mode: Passive vs. Active vs. No Inoculation) × 2 (Deepfake Attack: Pro-Attitudinal vs. Counter-Attitudinal) between-subjects experiment. Results show that inoculation increases deepfake awareness, intention to debunk deepfakes, and information-seeking behaviors, while reducing the perceived credibility of deepfake messages. However, exposure to counter-attitudinal deepfakes led to greater agreement with embedded disinformation.