The Automated Administration: Governance of ADM in the public sector

How can good public governance be achieved despite challenges associated with increased automation and AI systems in public decision-making? This is the overarching question that brings together researchers from Helsinki, Uppsala, and Södertörn universities, led by Stefan Larsson (Lund). Combining theoretical, empirical and legal approaches, this four-year multidisciplinary project studies and analyses Sweden and Finland, two countries with similar administrative foundations as well as forerunners in digitalisation.

The project is unique in that it combines both legal analysis, mapping of ADM system implementation, and citizens’ perceptions from a comparative perspective. This allows the development of empirically grounded and theoretically enhanced recommendations for future ADM developments that take into consideration current ethical challenges as well as the view of the affected citizens. The project will develop cutting-edge knowledge on how ADM systems are designed and what this means for the explainability of the decisions they produce, and what citizens perceive as meaningful information on automated decisions.

The project is tightly linked to vibrant Nordic research networks and includes reference group members from key national public authorities in Finland and Sweden. The research project will further deepen multidisciplinary cooperation between legal researchers and social scientists.

The project is divided into following working packages (WP):
WP1. Principled AI: from ethical principles to normative imaginaries (led by Stefan Larsson and Jockum Hildén)
How is the European principled approach to AI governance understood in the Nordic public administration and what are the consequences for ADM deployment?
WP2. Empirical: ADM strategy, implementation and supervision (led by Stefan Larsson, Riikka Koulu and Anne Kaun)
WP2a The public authority perspective
How are Nordic principles of good public governance taken into consideration in the practical implementation of AI systems, the formulation of AI strategy and the supervision of public authorities that use ADM?
WP2b The citizen perspective
How do citizens use, perceive and trust ADM systems when dealing with public authorities?
WP3. Legal analysis of ADM in Sweden and Finland (led by Jonas Ledendal and Riikka Koulu)
What safeguards to protect citizens’ rights, freedoms and legitimate interests with relation to AI and ADM are in place or currently developed in the Finnish and Swedish legal systems?

Combining theoretical, empirical and legal approaches, this four-year multidisciplinary research programme studies and analyses how good public governance can be achieved despite challenges associated with increased automation and AI systems in public decision-making. The programme focuses on Sweden and Finland, two countries with similar administrative foundations as well as forerunners in digitalisation. Led by the AI-focused socio-legal scholar Stefan Larsson, the programme spans the universities of Lund, Helsinki, Uppsala, and Södertörn, is tightly linked to vibrant Nordic research networks and includes reference group members from key national public authorities in Finland and Sweden.
The project is funded by the Future Challenges in the Nordics Research Programme (https://futurenordics.org/projects).