Project B06: Ethics and normativity of explainable AI

Explanations for decisions made by Artificial Intelligence are not always beneficial or desirable. For this reason, the co-construction of explanations has to be considered from an ethical point of view. Otherwise, explanations could be used to manipulate users or lead to unintended harms, says project leader Prof. Dr. Suzana Alpsancar.

During the first funding period, the project team from the fields of philosophy and media studies determined that there are a variety of normative reasons for explainable artificial intelligence (XAI). In the second funding phase, the project is investigating the organizational contexts in which XAI is used and how explainable AI can convey epistemic and political virtues in these contexts.

The overall aim of the project is to create a greater sensitivity to social contexts in the research of Artificial Intelligence. To this end, the project members will identify aspects that cannot be solved technically, but need to be addressed at a social or legal level.

 

Research areas: Philosophy, Media studies

Project Leaders

Prof. Dr. Suzana Alpsancar

More about the person

Prof. Dr. Tobias Matzner

More about the person

Staff

Violeta Milicevic

More about the person

Dr. Wessel Reijers

More about the person

Support Staff

Frederik Schauerte, Paderborn University

Former Members

Dr. Martina Philippi, Research associate

Pub­lic­a­tions



Civic Vice in Digital Governance

W. Reijers, in: Public Governance and Emerging Technologies, Springer Nature Switzerland, Cham, 2025.


Introduction to the Ethics of Emerging Technologies

W. Reijers, M. Thomas Young, M. Coeckelbergh, Introduction to the Ethics of Emerging Technologies, 2025.


How to Govern the Confidence Machine?

P. de Filippi, M. Mannan, W. Reijers, Regulation & Governance (2025).


Healthy Distrust in AI systems

B. Paaßen, S. Alpsancar, T. Matzner, I. Scharlau, ArXiv (2025).



Explainability and AI Governance

W. Reijers, T. Matzner, S. Alpsancar, in: M. Farina, X. Yu, J. Chen (Eds.), Digital Development. Technology, Ethics and Governance, Routledge, New York, 2025.


Warum und wozu erklärbare KI? Über die Verschiedenheit dreier paradigmatischer Zwecksetzungen

S. Alpsancar, in: R. Adolphi, S. Alpsancar, S. Hahn, M. Kettner (Eds.), Philosophische Digitalisierungsforschung  Verantwortung, Verständigung, Vernunft, Macht, transcript, Bielefeld, 2024, pp. 55–113.


AI explainability, temporality, and civic virtue

W. Reijers, T. Matzner, S. Alpsancar, M. Philippi, in: Smart Ethics in the Digital World: Proceedings of the ETHICOMP 2024. 21th International Conference on the Ethical and Social Impacts of ICT. Universidad de La Rioja, 2024., Longrono, 2024.


Unpacking the purposes of explainable AI

S. Alpsancar, T. Matzner, M. Philippi, in: Smart Ethics in the Digital World: Proceedings of the ETHICOMP 2024. 21th International Conference on the Ethical and Social Impacts of ICT, Universidad de La Rioja, 2024, pp. 31–35.







Interdisciplinary challenges for XAI ethics and the potential of the phenomenological approach. Gastvortrag am Eindhoven Center for the Philosophy of AI (ECPAI), TU Eindhoven, 25. Juni 2024.

M. Philippi, Interdisciplinary Challenges for XAI Ethics and the Potential of the Phenomenological Approach. Gastvortrag Am Eindhoven Center for the Philosophy of AI (ECPAI), TU Eindhoven, 25. Juni 2024., 2024.


(X)AI ethics for decision support systems. Gastvortrag am DLR Institut für den Schutz maritimer Infrastrukturen, Bremerhaven, 14. November 2024.

M. Philippi, (X)AI Ethics for Decision Support Systems. Gastvortrag Am DLR Institut Für Den Schutz Maritimer Infrastrukturen, Bremerhaven, 14. November 2024., 2024.




Blockchain Governance

W. Reijers, P. de Filippi, M. Mannan, Blockchain Governance, 2024.


What is AI Ethics? Ethics as means of self-regulation and the need for critical reflection

S. Alpsancar, in: International Conference on Computer Ethics 2023, 2023, pp. 1--17.



Technology and Civic Virtue

W. Reijers, Philosophy & Technology 36 (2023).


Show all publications