The article addresses the introduction and operation of devices supported by artificial intelligence (AI), in public action. In recent years, States and governments—including France—have increasingly used this type of instrument in a wide variety of areas. The use of algorithmic decision-making tools by public authorities is likely to have profound structural effects on the functioning of the State and its relations with the public. Based on this hypothesis, this article aims to question the transformative potential of innovations linked to algorithms and AI for various professions in public administration. The article examines the institutional frameworks and actors currently involved in the main projects of automation of State action, the legitimizing discourses, but also the practices of the concerned administrations. https://journals.openedition.org/socio/17434