Sometimes it's OK: How Citizens Understand and Experience the Use of Generative AI in Elections
Year: 2025
Type: preprint
Study design: Observational Study
Abstract: Generative AI (GenAI) presents opportunities and raises concerns about its role in creating misleading political content, threatening electoral integrity. Citizens in Germany, the Netherlands, and Poland (N = 2,894) were surveyed after the 2024 EU Parliament elections to examine encounters with AI-generated disinformation by threat type and modality. Findings revealed frequent perceived encounters with AI-generated text and disinformation targeting politicians, and showed heightened concern about threats posed by GenAI videos. Citizens accepted politicians’ use of GenAI to create disinformation on salient issues that aligned with prior beliefs. These insights may inform electoral and campaign policies and democratic resilience in the AI era.
Authors: Aqsa Farooq, Marieke van Hoof, Claes de Vreese
Language: English
Subfield: Sociology and Political Science
Field: Social Sciences
Domain: Social Sciences
Open Access status: gold
OpenAlex ID: W4413425973