Paper Type
Short
Paper Number
PACIS2025-1430
Description
The 2024 United States presidential election campaigns saw a proliferation of political audio-visual deepfakes featuring the two leading candidates. Although many deepfakes were not necessarily intended to trick voters, deepfake disinformation is a growing trend. Previous studies on fake news show that political partisanship significantly shapes beliefs and influences users' viewing and sharing intentions. In this study, we surveyed 400 U.S. voters to examine the role partisanship plays in how voters perceive and detect political audio-visual deepfakes. We find that deepfakes are highly effective vehicles for deception, and partisanship wields a statistically significant but limited influence on deepfake detection. Only progressives were statistically more likely to correctly detect a political deepfake among different ideological groups. Crucially, however, unlike the fake news phenomenon, we did not find that partisanship leads to a greater likelihood of detecting deepfakes that are discordant with one's political ideology and vice versa.
Recommended Citation
Kumar, Animesh; Burtscher, Christoph; and Eckhardt, Andreas, "Role of Partisanship in Detection of Political Audio-visual Deepfakes: A Study on the 2024 U.S. Presidential Election" (2025). PACIS 2025 Proceedings. 7.
https://aisel.aisnet.org/pacis2025/security/security/7
Role of Partisanship in Detection of Political Audio-visual Deepfakes: A Study on the 2024 U.S. Presidential Election
The 2024 United States presidential election campaigns saw a proliferation of political audio-visual deepfakes featuring the two leading candidates. Although many deepfakes were not necessarily intended to trick voters, deepfake disinformation is a growing trend. Previous studies on fake news show that political partisanship significantly shapes beliefs and influences users' viewing and sharing intentions. In this study, we surveyed 400 U.S. voters to examine the role partisanship plays in how voters perceive and detect political audio-visual deepfakes. We find that deepfakes are highly effective vehicles for deception, and partisanship wields a statistically significant but limited influence on deepfake detection. Only progressives were statistically more likely to correctly detect a political deepfake among different ideological groups. Crucially, however, unlike the fake news phenomenon, we did not find that partisanship leads to a greater likelihood of detecting deepfakes that are discordant with one's political ideology and vice versa.
Comments
Security