The use of AI technologies in civil service provision fundamentally alters citizen-state interactions, introducing new risks and design challenges to ensure safe, inclusive, and just applications that do not discriminate against vulnerable citizens. The VALUE-AI project addresses these challenges by developing a novel theory of value-based risk governance for AI technologies in civil service provision through collaborative research with municipal AI experts-practitioners. Using an innovative mixed-methods approach, the VALUE-AI research consortium will conduct co-creative, design-based research that synergistically links scholarship and practice to address the pressing, multi-dimensional issue of AI-related risk in civil service provision. A systematic, multi-disciplinary synthesis of AI risk governance scholarship will highlight ethical challenges and high-risk areas of AI applications in public administration. In-depth qualitative analysis will reveal how AI-related risks for vulnerable citizens are identified, deliberated, and addressed in the strategic and operational design of AI-based solutions in civil service provision. Together, this novel empirical evidence will allow us to derive an innovative and comprehensive value-based AI risk governance framework that advances public administration scholarship and offers evidence-based best practices and policy advice. With these insights, practitioners can create citizen-centric, trustworthy, and value-based AI applications that benefit both citizens and society.
