Abstract
Around the turn of the century, government digitalization mostly consisted of the automation of relatively simple bureaucratic procedures and the digitalization of organizations’ client databases. In this article, we argue that the ‘system-level bureaucracy’ has been surpassed by the emergence of information infrastructures, in which (big) data is shared among a wide variety of organizations, and the increased use of machine learning algorithms to assist administrative decision-making. We call this the ‘data-bureaucracy’. Rather than mere technical innovations developed to improve government efficiency, these developments have profound consequences for the way government organizations use public and private data, organize decision-making processes, and can be held accountable for their actions and decisions by citizens. We speak of the emergence of a ‘coding elite’ – data professionals that design concrete AI-applications and, in doing so, make (implicit) trade-offs between relevant public values beyond political and public scrutiny. In order to recover public value deliberation in algorithmic governance, we argue for the importance of inclusive design processes of AI-applications and develop a concrete framework for realizing such processes (‘inclusive AI’).
Original language | Dutch |
---|---|
Article number | BENM 2024/3 |
Pages (from-to) | 278-293 |
Journal | Beleid & Maatschappij |
Volume | 51 |
Issue number | 3 |
DOIs | |
Publication status | Published - 1 Dec 2024 |
Research programs
- SAI 2005-04 MSS