Real or Generated? Individual, Collective, and Expert Intelligence in Detection of AI-generated Media

This thesis investigates how individuals, online communities, and experts detect AI-manipulated images and videos. Using Reddit discourse analysis and a quantitative experiment, it compares detection accuracy and visual cues across these groups, with a focus on how collective intelligence shapes outcomes. Findings will inform the design of media verification tools.

This thesis is currently reserved.
To join the waiting list, please contact Liudmila Zavolokina