AUTO-WELF: Automating Welfare – Algorithmic Infrastructures for Human Flourishing in Europe

enlarge the image: Welfare State Automatization
Getty Images

How is welfare automated across European countries and across different welfare regimes?

The project examines the effects of algorithms and artificial intelligence on the well-being and prosperity of European citizens in eight countries. It investigates the extensive implementation of automated decision-making (ADM) in the welfare sector across Europe providing a comparative analysis. AUTO-WELF foregrounds the perspective of people implicated in the automation process, including infrastructural engineers and designers, case workers and citizens. In doing so, the project implements a multimethod, interdisciplinary and cross-country comparative approach, to study two domains related to broader visions of automated welfare: a) automation of core welfare, and b) automation of communal welfare infrastructures and services. These domains operate from individual and aggregated levels of data, respectively, and thus enable us to compare automated welfare through a maximum variation design. The project addresses three important goals: first, to develop comparative baselines for automating welfare across the EU; second, to develop a theoretical contribution to conceptualise the automation welfare for the benefit of people, and thirdly, it produces guidelines for policymakers.

In order to achieve these research objectives (aims), the project adopts a cross-country comparative design acknowledging the importance of contextualising ADM across different countries and welfare state models to highlight the cultural, economic, social and political specificities of ADM systems. Therefore, the study will provide an eight-country comparison (Sweden, Denmark, Estonia, Poland, Germany, Austria, Portugal and Italy), focusing firstly on the relationship between welfare-state regimes and automated decision-making, and, secondly, on the prospects and pitfalls of welfare automation across the individual and communal levels of human flourishing. Based on this, the two domains of core welfare services and communal welfare infrastructures will be explored by implementing a mixed-methods approach that includes citizen data journeys mapping per country, case studies on organisational ethnographies and expert interviews, and mind scripting and vision workshops with citizens. Hence, the comparative empirical results contribute to theory building on new algorithmic public services and to respective policy recommendations.

The project develops a strong evidence-based impact for different stakeholders with the following five results:

First: Empirically develops the understanding of ADM by offering a comparative analysis to advance welfare provision in Europe

Second: Conceptually advances the understanding of changes in welfare provision with and through digital data-driven technologies,

Third: Engages with the implications for the people affected by automated decision-making across welfare regimes.

Fourth: Informs solutions to emerging challenges of automated decision-making in the public sector by identifying best and worst practice examples

Fifth: Develops a multi-method and interdisciplinary approach that combines organisational fieldwork, mapping of citizens’ data journeys and mind scripting and vision workshops and brings theoretical discussions of algorithmic governance and digital culture to the field of public welfare research.

Head of Project

Prof. Dr. Christian Pentzold

Prof. Dr. Christian Pentzold


Media and Communications
Nikolaistraße 27-29, Room 5.05
04109 Leipzig

Phone: +49 341 97-35701
Fax: +49 341 97-35749


  • Project Leader: Anne Kaun, Södertörn University, Department for Culture and Education, Sweden
  • Stine Lomborg, University of Copenhagen, Department of Communication, Denmark
  • Christian Pentzold, Leipzig University, Institute for Communication and Media Studies, Germany
  • Karolina Sztandar-Sztanderska, Institute of Philosophy and Sociology Polish Academy of Science, Poland
  • Doris Allhutter, Austrian Academy of Sciences, Institute of Technology Assessment, Austria


Cooperation Partners

  • Matthias Spielkamp, Algorithm Watch
  • Birgitte Kofod Olsen, DataEthics
  • Tanja Mally, – Plattform Grundrechtspolitik
  • Pernille Boye Kock, The Danish Institute for Human Rights
  • Ana Jorge, Lusófona University
  • Alice Mattoni, University of Bologna
  • Brigitte Alfter, Arena for Journalism in Europe / DataHarvest
  • Daniel Neugebauer, Haus der Kulturen der Welt
  • Veronika Liebl, Ars Electronica Linz GmbH & Co KG

October 2022 to October 2025 (36 months)

The project is funded by CHANSE and BMBF, see here the Funding organisations