Automated Administrative Decision-Making: What is the Black Box Hiding?

Abstract

The exploration of the “black box” phenomenon underscores opacity challenges in automated administrative decision-making systems, prompting a discussion on the paradox of transparency. Advocating for the concept of “qualified transparency”, the article aims to navigate the delicate balance between understanding and safeguarding sensitive information. Ethical imperatives, including respect for human autonomy, harm prevention, fairness, and explicability, are considered, culminating in recommendations for human participation, ethicality or accountability by design considerations, and the implementation of regulatory sandboxes to test such models prior to broad integration. Ultimately, the article advocates for a comprehensive discourse on transitioning from a human-centric to an automated public administration model, acknowledging the complexity and potential risks involved.