[Air-L] Publication: Repairing AI

[Schaefer, M.T. (Mirko)]

Dear colleagues,
Our article, “Repairing AI: How Breakages of Algorithmic Systems Reveal Repair Efforts and Possibilities for Renewal,” has recently been published in Critical AI.
In the article, we draw on broken world theory to examine what happens when algorithmic systems fail, particularly when they violate social norms and public values. We analyse these "breakages" as moments that expose underlying assumptions, institutional arrangements, and accountability gaps.
Breakages often trigger repair efforts. We show how these processes of repair can lead to renewed responsibilities, strengthened checks and balances, and, in some cases, institutional change. By focusing on breakages and repair, the article offers a perspective on AI governance that centres maintenance, oversight, and democratic accountability.
The full article is available here (open access): https://doi.org/10.1215/2834703X-12095964
This article originated in the context of the Repair Project at Helsinki University (directed by Minna Ruckenstein). https://www.repair-research.fi/home-en

We are very much interested in the work of colleagues who are also investigating the "behind the scenes efforts" of repairing, contesting, and holding accountable of algorithmic systems. Please contact us for comparing notes and methods of inquiry.

Best wishes,
mirko

Mirko Tobias Schäfer | Associate Professor for AI, Data & Society|
Sciences Lead at Data School |
Department of Information & Computing Sciences |
Utrecht University | ICS | Princetonplein 5 | 3584 CC Utrecht |
 M: m.t.schaefer at uu.nl<mailto:m.t.schaefer at uu.nl> | W: https://www.uu.nl/staff/MTSchaefer