From Risk Regulation to Risk Governance:

Human rights law, the GDPR, the LED, and the AI Act, establish obligations for addressing discrimination in algorithmic profiling but leaves room regarding the operationalization of these obligations. Against this backdrop, the Dutch Ministry of Internal Affairs, Directorate of Digital Society (DDS), is developing the Algoritmekader, a framework to provide further guidance to public organizations regarding the responsible and non-discriminatory use of algorithmic systems. This paper examines how DDS approaches developing the Algoritmekader to address discrimination in algorithmic profiling and assesses the suitability of their approach for managing this systemic risk. Using empirical legal methods, the approach taken by DDS is analyzed through the lens of the systemic risk governance principle—communication and inclusion, integration, and reflection. The findings highlight the importance of integrating citizen perspectives in systemic risk governance, combining bottom-up & top-down coordination, strategically engaging stakeholders through targeted consultations, and embedding formal reflections to enhance institutional learning.