Journal of Risk Management in Financial Institutions
Subramanian (Subbu) Narayanaswamy, Executive Director, Wells Fargo Bank, US
Submission deadline: 15th January 2027
Credit risk management is undergoing its most consequential transformation in decades. Foundation models, alternative data, real-time decisioning engines, and, increasingly, agentic AI systems are changing how institutions assess creditworthiness, price risk, and allocate capital, while fintech lenders and embedded finance platforms blur the lines between origination, servicing, and risk oversight. The opportunity and the fragility are happening simultaneously: AI can widen credit access and improve risk differentiation, but it can also encode bias at scale, obscure accountability, and create model dependencies that are poorly understood until they fail.
A second transformation has received far less attention. AI is becoming a driver of credit risk in its own right. The capital expenditure cycle financing AI infrastructure is creating increasingly concentrated exposures across banks, private credit providers and capital markets, some of which are difficult to see through conventional measures of direct lending. AI-driven changes in employment and income stability could affect consumer and small business portfolios in ways that models calibrated on an earlier labour market may not capture. The industry has spent several years asking how AI changes credit decisions; the question for the next cycle is how AI changes credit.
Regulation is moving unevenly. In the European Union, obligations governing high-risk AI systems, including those used to assess the creditworthiness of natural persons, became applicable in August 2026. In the United States, revised interagency model risk management guidance placed generative and agentic AI outside its scope, leaving institutions to work out how existing expectations apply. Institutions are deploying some of their least understood technology in an area where the formal governance framework remains unsettled.
This special issue seeks contributions that assess where this is all heading, not only with where it stands today. We are particularly interested in work grounded in production experience, supervisory practice or original data. Practice articles, case studies and applied research are all welcome, from both academic and practitioner authors.