Towards fair AI in Estonia’s public service: discussing and disseminating bias prevention in automated decision-making

Abstract Estonia is recognized as a pioneer state in digitalizing its public services, particularly through its integration of artificial intelligence (AI) systems. However, the lack of tailored strategies and tools to address biases in AI-based decision-making in Estonian public services processes poses significant ethical challenges. This study explores the strategies and tools available to identify and prevent bias in automated decision-making (ADM) processes and ways of their dissemination. It draws from the European Commission-funded EquiTech project, which seeks to ensure fairness in algorithmic decision-making in public services. Employing mainly qualitative methodology, the study draws on literature review, document analysis and discussions conducted in the framework of the EquiTech project. Through this exploration and dissemination, the study contributes to the conceptualization of tailored strategies and tools in Estonia, ensuring fairness in ADM processes within public services, promoting equitable digital governance and mitigating discriminatory outcomes.