Generative AI Adoption in Banking Institutions: A Design Science Approach to Addressing Trust, Data Sovereignty, Cost and Workforce Resistance Barriers

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

Despite significant pilot activity in Generative AI across financial institutions (78% running active pilots), only a fraction transition these projects into operational production. While commercial maturity exists for underlying technologies such as retrieval-augmented generation and private endpoints, significant organizational barriers prevent deployment. This paper employs a Design Science Research (DSR) methodology to address four interconnected domains of organizational resistance: (1) Trust in model output (explainability, bias, hallucination), (2) Data Sovereignty and Regulatory Compliance (GDPR, EU AI Act, Basel rules), (3) Economic Viability (TCO, delayed returns, permanent control overhead), and (4) Workforce Resistance (change management and turnover). Rather than treating these barriers in isolation, this study formulates an integrated governance framework to evaluate their interactive effects, providing actionable decision support for executive leadership in banking institutions.