EUVIMEDEUVIMED entdecken
Systematische ÜbersichtOpen Access · CC BY 4.0

Explainable AI in Small-Business Credit Scoring: A Systematic Literature Review of Transparency, Model Governance, and U.S. Fair-Lending Gaps in Hybrid-Cloud Environments

Amit Butail, Anamika, Dr. Kritika Rana

Computer and Decision Making: An International Journal · 09.11.2026 · 3 (8S) · 775-806

DOI: 10.1922/comdem.v3i8s.19226 ↗

Kurz gefasst

Systematische Übersicht in der Zeitschrift Computer and Decision Making: An International Journal, online erschienen am 09.11.2026. Verfasst von Amit Butail und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Artificial intelligence (AI) is becoming more prevalent in small-business credit scoring, and the model must be accurate, explainable, fair, operationally accountable, and comply with U.S. lending regulations. This systematic literature review is a synthesis of the evidence pertaining to machine-learning credit assessment, explainable AI (XAI), algorithmic fairness, hybrid-cloud governance, and regulatory compliance. The review identified 287 records, with 55 studies included in the qualitative synthesis, according to PRISMA 2020 reporting principles. The synthesis differentiates 14 studies that provide direct evidence of SME credit from 41 complementary studies that cover more general credit and financial-governance contexts. Ensemble learning, gradient boosting, deep learning, transfer learning, SHAP, LIME and counterfactual explanations are explored in the literature. But it is important to note that there are differences in borrower populations, data sets, and modelling approaches that make direct comparisons difficult. Evidence of explanation fidelity, stability, subgroup fairness and cross-cloud accountability is still limited, especially in terms of decision reconstruction and auditability in hybrid-cloud environments. There are fair-lending and adverse-action requirements in the Equal Credit Opportunity Act and Regulation B, and supervisory guidance and voluntary AI frameworks offer additional governance principles. Six research gaps were identified: predictive validation, explanation reliability, fairness assessment, hybrid-cloud controls, lifecycle governance, and regulatory traceability. The synthesis supports a five-layer conceptual architecture integrating data governance, predictive modelling, explainability, fairness evaluation, and regulatory oversight with cross-cutting security controls. Study-level quality appraisal and screening documentation remain incompletely verified. The proposed architecture requires empirical validation in representative U.S. small-business lending environments.

Zitieren

APA

Butail, A., Anamika, & Rana, D. K. (2026). Explainable AI in Small-Business Credit Scoring: A Systematic Literature Review of Transparency, Model Governance, and U.S. Fair-Lending Gaps in Hybrid-Cloud Environments. Computer and Decision Making: An International Journal, 3(8S), 775-806. https://doi.org/10.1922/comdem.v3i8s.19226

Vancouver

Butail A, Anamika, Rana DK. Explainable AI in Small-Business Credit Scoring: A Systematic Literature Review of Transparency, Model Governance, and U.S. Fair-Lending Gaps in Hybrid-Cloud Environments. Computer and Decision Making: An International Journal. 2026;3(8S):775-806. doi:10.1922/comdem.v3i8s.19226

BibTeX
@article{Butail2026,
  author = {Butail, Amit and Anamika and Rana, Dr. Kritika},
  title = {Explainable AI in Small-Business Credit Scoring: A Systematic Literature Review of Transparency, Model Governance, and U.S. Fair-Lending Gaps in Hybrid-Cloud Environments},
  journal = {Computer and Decision Making: An International Journal},
  year = {2026},
  volume = {3},
  number = {8S},
  pages = {775--806},
  doi = {10.1922/comdem.v3i8s.19226}
}
Kostenlos starten →