A Governance-Oriented Information Systems Framework for Explainable and Trustworthy Deep Learning in Organizational Decision Support
Sage Open · 2026-10 · 16 (4)
DOI: 10.1177/21582440261503464 ↗Kurz gefasst
Fachartikel in der Zeitschrift Sage Open, online erschienen am 03.11.2026. Verfasst von Abdalilah Alhalangy und 1 weiteren Autor:innen. Die Arbeit nennt 57 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).
Abstract (Originalsprache)
Organizations are increasingly embedding deep learning outputs into operational decision-support systems for prioritization, classification, screening, and recommendation. This shift creates a governance challenge for information systems: although model outputs may improve decision quality, they can also weaken reviewability, traceability, and organizational accountability when deployed without explicit oversight and control. This paper develops a governance-oriented information systems framework for integrating deep learning into organizational decision support in ways that remain explainable, governable, and operationally reliable over time. Using a design science approach, the study derives requirements from prior research on information systems, explainable and trustworthy AI, human oversight, decision support, and lifecycle risk management. These requirements are translated into meta-requirements, design principles, and a layered architecture, and are evaluated through structured analytical validation, scenario-based stress testing, and retrospective analysis of documented AI governance failures. The resulting framework organizes trustworthy integration across six interconnected layers: data governance, model validation and readiness, decision workflow integration, human oversight and escalation, risk and control mechanisms, and traceability with continuous monitoring. The study contributes to organizational information systems research by showing how explainability, oversight, and accountability can be embedded within decision-support arrangements rather than treated as external ethical aspirations. The framework offers practical guidance for designers, governance teams, and decision-support stakeholders operating in high-impact settings where reviewability and post-deployment oversight are essential.
Zitieren
APA
Alhalangy, A., & Alkhamis, S. A. (2026). A Governance-Oriented Information Systems Framework for Explainable and Trustworthy Deep Learning in Organizational Decision Support. Sage Open, 16(4). https://doi.org/10.1177/21582440261503464
Vancouver
Alhalangy A, Alkhamis SA. A Governance-Oriented Information Systems Framework for Explainable and Trustworthy Deep Learning in Organizational Decision Support. Sage Open. 2026;16(4). doi:10.1177/21582440261503464
BibTeX
@article{Alhalangy2026,
author = {Alhalangy, Abdalilah and Alkhamis, Saleh Abdulrahman},
title = {A Governance-Oriented Information Systems Framework for Explainable and Trustworthy Deep Learning in Organizational Decision Support},
journal = {Sage Open},
year = {2026},
volume = {16},
number = {4},
doi = {10.1177/21582440261503464}
}
EUVIMED