AI and transparency

Aula, Ville; and Erkkilä, Tero (2024) AI and transparency. In: Handbook on Public Policy and Artificial Intelligence. Handbooks of Research on Public Policy series . Edward Elgar, Cheltenham, UK, 170 - 180. ISBN 9781803922164
Copy

This chapter critically explores how the concept of transparency is used in the emerging scholarly and policy debates on AI systems, focusing on its political character. The concept of transparency has evolved over time, containing democratic and economic connotations, but is now gaining new technical interpretations in the context of AI systems. This conceptual shift is important, because the emerging sociotechnical understanding of algorithmic transparency foregrounds specific framings of policy problems while deprioritizing others. As a political concept algorithmic transparency hence carries the potential to be instrumentalized for promoting technical and ethical solutions to AI instead of considering its broader democratic and economic aspects. We conclude that current scholarly and policy debates on algorithmic transparency fall short on establishing actual institutional arrangements through which civil society actors and key stakeholders could control the use of algorithmic and AI applications.

picture_as_pdf

picture_as_pdf
subject
Accepted Version

Download

Atom BibTeX OpenURL ContextObject in Span OpenURL ContextObject Dublin Core MPEG-21 DIDL Data Cite XML EndNote HTML Citation METS MODS RIOXX2 XML Reference Manager Refer ASCII Citation
Export

Downloads