Fighting fake news with fake faces: the effects of deepfake self-debunking on misinformation correction

Misinformation remains a persistent challenge, particularly when spread by influential figures. This study introduces deepfake self-debunking – a novel misinformation correction method that uses deepfake technology to depict a misinformation source retracting their own false claims. An online experiment (N = 1,346) tested the effectiveness of three deepfake self-debunking styles – serious, purely humorous, and satirical – compared to a control condition. Results showed that all deepfake conditions reduced misbeliefs. However, no differences emerged for policy support. Consistent with the Disposition Theory of Humor, the purely humorous video was more effective for misinformation correction among supporters of the misinformation source than the satirical video. Counterarguing mediated these effects, providing insight into the psychological mechanisms. Findings suggest that pure humor may be the optimal strategy for engaging resistant audiences, offering practical and ethical considerations for using deepfake technology in misinformation correction.