The conference brought together participants from academia, public administration and the economy to discuss current developments and challenges in digital public administration. The Match’In project was represented as well. In his workshop, Dr Christian Sauer gave a detailed presentation on the project, focusing on the challenges posed by algorithmic bias and on approaches and methods for avoiding it, both within the context of the project as well as in the use of AI systems in general. The project relies on case-based reasoning to ensure the traceability of algorithmic recommendations and thus to avoid algorithmic bias.
The following discussion focused, amongst other things, on the responsibility for unexplainable AI-decisions and the question of whether public administrations currently have sufficient capacities to regularly check AI-systems for potential algorithmic bias. The discussion and suggestions provided valuable input for the further development of Match’In 2.0.
The participation in the N3GZ Spring Conference has once again demonstrated how important the exchange between research, public administration and practical application is. The insights gained there are fed directly into the further development of Match’In, with the aim of designing AI-supported solutions for public administration in a way that is both responsible and transparent.
