Authors: M Wack, DA Parry
Year: 2025
Published in: International Journal of Communication, 2025 - ijoc.org
Institution: University of Zurich
Research Area: Generative AI, Disinformation, Political Communication, Ethnic Targeting
Discipline: Communication, Artificial Intelligence, Political Science
The study finds that AI-generated political ads with coethnic avatars are more effective at mobilizing voter support and reducing skepticism, even when labeled as synthetic, with AI literacy playing a key role in identifying such content.
Methods: Survey experiment targeting voter responses to AI-generated political ads with varied presenter ethnicities, including analysis of AI literacy versus digital literacy.
Key Findings: Effectiveness of coethnic versus out-group ethnic AI-generated avatars in mobilizing voter support and the role of AI literacy in detecting synthetic content.
Citations: 1