Short abstract:

AI-driven automation is expected to transform welfare benefits. Our study combines frameworks of boundary work and ‘regimes of justification’ to examine the construction of legitimacy for the GenAI tool within public administration, offering crucial insights into the sustainability of AI automation.

Long abstract:

Internationally, public administration and services are being transformed by AI-driven automation processes, promising enhanced efficiency and service quality in welfare benefits. Our study combines theories of boundary work (Gieryn 1983, 1995), insights from the sociology of professions into professional life and struggles (Abbott 1988, 1995, 2005), and a framework of ‘regimes of justification’ (Boltanski and Thévenot 2006) to identify key actors, their practices, interpretations and claims surrounding a GenAI tool conceptualised as a boundary object that brings together a multidisciplinary innovation team, service design consultants, and other intra-organisational actors. Our methodology is rooted in a comprehensive year-long ethnographic study in a Finnish public sector institution involving extensive observation and interviews. We present our empirical findings as organized into sets of legitimisation frames that emerge, stabilise and underpin the construction of the GenAI tool’s legitimacy. These frames are analyzed as related to various forms of boundary work instrumental in legitimizing such tools within the public sector while engaging in negotiations of public values. Our contributions offer critical insights into the social sustainability of AI-driven automation within the Finnish public administration landscape.