Governing Algorithms: Transparency, Digitalization, and Risk in European Public Administration – A Comparative Study

This paper investigates the evolution of algorithmic governance in public administration by comparatively analyzing Estonia, the Netherlands, France, and Romania over the period 2020–2026. Grounded in the theoretical framework of digital governance and algorithmic accountability, the study operationalizes four core dimensions: digitalization (DI), AI integration (AIU), transparency (TI), and algorithmic risk (AR). Methodologically, the research employs a comparative and visual analytical approach, integrating clustered heatmaps, longitudinal trend analysis, and radar chart representations to capture both structural patterns and temporal dynamics. The results reveal a consistent alignment between high levels of digitalization and AI adoption with increased transparency and reduced algorithmic risk, supporting the hypothesis that governance capacity co-evolves with technological advancement. The analysis further identifies differentiated national trajectories: Estonia and the Netherlands exemplify consolidated models of algorithmic governance, France reflects an intermediate institutional adaptation, while Romania illustrates an asymmetrical development pattern characterized by accelerated digital uptake but comparatively weaker governance safeguards. The study argues that the effectiveness of algorithmic governance depends on the institutionalization of transparency and accountability mechanisms, emphasizing the need for coherent policy frameworks that integrate innovation with risk mitigation. These findings contribute to the broader discourse on responsible AI and public sector digital transformation.