The Deepfake Democracy Project is an interdisciplinary research initiative at Malmö University investigating how AI-generated video deepfakes threaten democratic resilience in European and Nordic contexts. Using a Vertically Integrated Project (VIP) model, we combine technical innovation, social science analysis, design solutions, and policy governance to develop evidence-based interventions that strengthen epistemic trust and democratic participation.
Our Research
The Deepfake Democracy Project operates as a long-term interdisciplinary research initiative at Malmö University addressing the question: How do video deepfakes deployed in electoral and political communication threaten democratic resilience in European and Nordic contexts, and how can interdisciplinary interventions mitigate these harms? The group’s research is structured around a Vertically Integrated Project (VIP) model that brings together students and faculty across five complementary research tracks, employing mixed methods that combine technical analysis, comparative case studies, content analysis, participatory design, and policy research.
Track A: AI Creation Tools Analysis – computer science researchers systematically document deepfake creation platforms, develop testing frameworks, and contribute to detection tools research.
Track B: Detection & Verification Systems – working at the intersection of computer science and engineering, the group benchmarks leading detection tools, identifies system vulnerabilities, and develops improved detection methodologies.
Track C: Democratic and Social Impact Analysis – researchers from global political studies, media studies, and criminology document deepfake incidents affecting electoral processes through comparative case studies while analysing media responses, public trust dynamics, and psychological impacts.
Track D: Design and Mitigation Solutions – interaction designers and product designers create user interface and user experience solutions for content verification systems.
Track E: Policy and Governance – researchers from global political studies, criminology, and legal studies analyse international deepfake legislation and develop policy recommendations for EU AI Act compliance and democratic resilience.
The group pursues an active dissemination strategy through multiple channels, plans to publish peer-reviewed journal articles on deepfakes and democratic governance, present findings at conferences, and host cross-Öresund workshops on democratic resilience.
The project maintains research partnerships across Scandinavia with the University of Copenhagen, Lund University's Raoul Wallenberg Institute, as well as with European organisations, including AI Sweden and Denmark AI.
