Governments are increasingly adopting algorithmic technologies to detect and prevent welfare fraud, marking a shift in the balance between providing care and exercising control. By repurposing data collected for benefits and services to identify fraudulent behavior, these technologies reshape the relationship between citizens and the state. While scandals surrounding algorithmic fraud detection in countries like Australia, Denmark, and the Netherlands have gained significant attention, comparative research that moves beyond single case studies remains scarce, particularly regarding how citizens perceive these systems and their societal implications.
This interdisciplinary and cross-national project examines the use of algorithmic fraud detection in welfare systems in Estonia, Poland, and Sweden. It aims to: (a) analyze the key discourses driving the adoption of these technologies, (b) explore citizens’ perceptions of algorithmic systems for fraud detection, and (c) comparatively investigate how these developments influence understandings of social rights and obligations. For this purpose the project combines analysis of publicly available documents, media articles, and expert interviews with focus groups to disentangle algorithmic welfare fraud detection, or what we call algorithmic forensics, as a socio-technical system that is formed by discourses, perceptions, and technological infrastructures.