Open Journal of Business and Management

Volume 13, Issue 4 (July 2025)

ISSN Print: 2329-3284   ISSN Online: 2329-3292

Google-based Impact Factor: 2.35  Citations  

Ethical AI in Banking: A Sectoral Literature Review with Case-Based Interventions

  XML Download Download as PDF (Size: 287KB)  PP. 2420-2430  
DOI: 10.4236/ojbm.2025.134125    4 Downloads   72 Views  
Author(s)

ABSTRACT

As artificial intelligence (AI) becomes more integrated into banking, concerns around bias, privacy, and governance are becoming increasingly urgent. This article explores the ethical risks of AI in financial services and reviews practical interventions that can help banks use AI responsibly. Drawing from literature and real-world examples from institutions like JP Morgan Chase, Wells Fargo, HSBC, Bank of America, Triodos Bank and Credit Union Australia, the study highlights five key strategies: implementing Explainable AI (XAI), using bias detection tools, forming ethics committees, educating customers, and aligning with global standards such as the EU AI Act, OECD AI Principles, and NIST AI RMF. The article evaluates tools like IBM’s AI Fairness 360, Google’s What-If Tool, and Fairkit-learn for detecting and mitigating bias in automated decisions. While challenges such as cost, technical complexity, and organizational resistance exist, especially for smaller banks, the findings show that ethical, transparent AI is achievable and scalable. The paper concludes that strong leadership, stakeholder engagement, and policy alignment are critical for building AI systems that are fair, trustworthy, and fit for the future of banking.

Share and Cite:

Smit, J. (2025) Ethical AI in Banking: A Sectoral Literature Review with Case-Based Interventions. Open Journal of Business and Management, 13, 2420-2430. doi: 10.4236/ojbm.2025.134125.

Cited by

No relevant information.

Copyright © 2025 by authors and Scientific Research Publishing Inc.

Creative Commons License

This work and the related PDF file are licensed under a Creative Commons Attribution 4.0 International License.