Measuring adherence to AI ethics: a methodology for assessing adherence to ethical principles in the use case of AI-enabled credit scoring application
AI and Ethics · 2025-04 · 5 (2) · 1291-1313
DOI: 10.1007/s43681-024-00468-9 ↗Kurz gefasst
Fachartikel in der Zeitschrift AI and Ethics, online erschienen am 15.04.2024. Verfasst von Maria Pokholkova und 3 weiteren Autor:innen. Die Arbeit wurde bisher 11-mal zitiert (Crossref) und nennt 69 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Fujitsu Limited, Technische Universität München.
Themen: Ethik in der Pflege
Abstract (Originalsprache)
This article discusses the critical need to find solutions for ethically assessing artificial intelligence systems, underlining the importance of ethical principles in designing, developing, and employing these systems to enhance their acceptance in society. In particular, measuring AI applications’ adherence to ethical principles is determined to be a major concern. This research proposes a methodology for measuring an application’s adherence to acknowledged ethical principles. The proposed concept is grounded in existing research on quantification, specifically, Expert Workshop, which serves as a foundation of this study. The suggested method is tested on the use case of AI-enabled Credit Scoring applications using the ethical principle of transparency as an example. AI development, AI Ethics, finance, and regulation experts were invited to a workshop. The study’s findings underscore the importance of ethical AI implementation and highlight benefits and limitations for measuring ethical adherence. A proposed methodology thus offers insights into a foundation for future AI ethics assessments within and outside the financial industry, promoting responsible AI practices and constructive dialogue.
Förderung
- Fujitsu Limited
- Technische Universität München
Zitieren
APA
Pokholkova, M., Boch, A., Hohma, E., & Lütge, C. (2025). Measuring adherence to AI ethics: a methodology for assessing adherence to ethical principles in the use case of AI-enabled credit scoring application. AI and Ethics, 5(2), 1291-1313. https://doi.org/10.1007/s43681-024-00468-9
Vancouver
Pokholkova M, Boch A, Hohma E, Lütge C. Measuring adherence to AI ethics: a methodology for assessing adherence to ethical principles in the use case of AI-enabled credit scoring application. AI and Ethics. 2025;5(2):1291-1313. doi:10.1007/s43681-024-00468-9
BibTeX
@article{Pokholkova2025,
author = {Pokholkova, Maria and Boch, Auxane and Hohma, Ellen and Lütge, Christoph},
title = {Measuring adherence to AI ethics: a methodology for assessing adherence to ethical principles in the use case of AI-enabled credit scoring application},
journal = {AI and Ethics},
year = {2025},
volume = {5},
number = {2},
pages = {1291--1313},
doi = {10.1007/s43681-024-00468-9}
}
EUVIMED