AI, state and bureaucratic reinvention: a theoretical model for new state capacities in the 21st century

This theoretical article examines the transformative impact of Artificial Intelligence (AI) on traditional state capacities, presenting an original conceptual model that repositions bureaucratic competencies in light of automation, predictive analytics, and machine learning. Drawing from interdisciplinary theories of institutional capacity, public innovation, and algorithmic governance, the study explores three interrelated dimensions: (i) the reconfiguration of public policy analysis through big data and algorithmic tools; (ii) the automation of implementation functions in state services; and (iii) the ethical, normative, and epistemological challenges involved in delegating decisions to AI systems. The proposed model reveals how AI reshapes bureaucratic roles, blurs the boundary between technical rationality and political discretion, and inaugurates new forms of technopolitical governance. It highlights emerging risks such as algorithmic opacity, reduced accountability, and erosion of discretionary autonomy. The study advocates for a new framework of public algorithmic governance rooted in transparency, explainability, and institutional safeguards. Far from proposing a deterministic perspective, the article invites critical reflection on how AI mediates power, redefines legitimacy, and generates both opportunities and dilemmas for 21st-century statecraft. It contributes to advancing theoretical debates on digital bureaucracy, algorithmic justice, and the future of democratic governance.