AUTHOR=Zenkl Thomas 

TITLE=The taming of sociodigital anticipations: AI in the digital welfare state

JOURNAL=Frontiers in Sociology

VOLUME=Volume 10 - 2025

YEAR=2025

URL=https://www.frontiersin.org/journals/sociology/articles/10.3389/fsoc.2025.1556675

DOI=10.3389/fsoc.2025.1556675

ISSN=2297-7775

ABSTRACT=Advanced algorithmic technologies and artificial intelligence (AI) are projected to profoundly impact public employment services (PES) and the delivery of labor market policies. Focusing on the perspectives of PES counselors in Austria, this study examines workers’ anticipations of future AI technologies. Based on face-to-face interviews (n = 23), it identifies and explores the concept of “taming” as a primary tactic for managing and appropriating uncertain AI futures within the limits of current institutional and bureaucratic structures. “Tamed” anticipations of advanced algorithms are rooted within challenging working conditions (insufficient resources and time for clients), reconfigurations of roles and agencies (administration of systems instead of supporting clients) and nested within transformations of techno-bureaucratic regimes (from street- over screen- to system-level bureaucracies), which they envision to rectify and repair. Taming is thus applied as a lens to observe practices of anticipation that navigate and contest precarities of daily work life beyond dichotomies of compliance and resistance. Despite the differences in tamed AI futures identified, the analysis shows corresponding properties: Tamed anticipations consolidate dependencies between human and machinic actors, demarcate what is (un-)tamable, and are inherently dialectical. Synthesizing structural problems and an imperative for human elements into futures that rely on advanced technologies, a truly “human” counseling situation is thought to be obtained through “machinic” means. To “tame” futures with and by anticipating certain AI technologies here means affirming ideas of a supposedly better working conditions (more humanity and margin of discretion) and regaining (or even improving) them through the computational means that are considered responsible for their loss.