مجلة التنمية وإدارة الموارد البشرية

Algorithmic Governance In The Digital State: Machine Learning, Bias, And Accountability In Public Administration

Auteurs : Benrahou Sihem .

Résumé

The integration of machine learning and artificial intelligence into state bureaucracies represents a fundamental transformation in governance practices, introducing what scholars term "algorithmic governance"—the delegation of administrative decision-making to automated systems. This article examines how states deploy machine learning in welfare targeting, policing, border control, and smart-city management, with particular attention to issues of bias, discrimination, and accountability. Drawing on comparative case analysis across multiple states and recent empirical evidence, this study reveals that while algorithmic systems promise efficiency gains, they simultaneously introduce complex challenges related to algorithmic discrimination, accountability deficits, and democratic legitimacy. Key findings indicate that algorithmic governance disproportionately affects marginalized communities, with documented cases of discrimination in welfare fraud detection, predictive policing, and border surveillance. The article proposes a comprehensive governance framework emphasizing transparency, human oversight, bias auditing, and legal accountability mechanisms to ensure algorithmic systems align with democratic principles and human rights standards.

Mots clés

algorithmic governance ; machine learning ; bias ; discrimination ; accountability ; welfare targeting ; predictive policing ; smart cities, ; public administration