Financial Risk Management in the Digital Era: A Systematic Review of Emerging Models, Technologies, and Practices
Abstract
Background. Digital technologies are transforming how financial institutions and corporations identify, measure and manage risk. Artificial intelligence (AI), big data, cloud computing, distributed ledgers and regulatory technology have produced a rapidly growing but fragmented literature.
Objective. This systematic review synthesises research on emerging models, technologies and practices in financial risk management, identifies their reported benefits and limitations, and sets out an agenda for research and practice.
Methods. The review followed the PRISMA 2020 guidelines. Scopus, Web of Science, IEEE Xplore and SSRN were searched for peer-reviewed studies and working papers published between January 2010 and December 2024, supplemented by reports from regulators and international bodies. Studies were screened against predefined eligibility criteria, appraised for quality and synthesised thematically.
Results. Six themes emerged: machine-learning models for credit risk; real-time analytics for fraud and anti-money laundering; data-driven approaches to market and liquidity risk; operational and cyber risk analytics; regulatory and supervisory technology; and climate and ESG risk analytics. Across themes, machine-learning models and alternative data generally improved predictive accuracy over traditional statistical models, although gains varied widely with data structure and evaluation design. Adoption is advancing quickly: a 2024 survey found that 75 percent of UK financial firms were already using AI. In India, Unified Payments Interface (UPI) transactions grew from about 84 billion in 2022–23 to 131 billion in 2023–24, while frauds reported by banks rose from 13,564 to 36,075 cases. The literature consistently identified explainability, fairness, data quality, model risk, cybersecurity and skills as constraints.
Conclusions. Digital technologies offer substantial gains in the speed and accuracy of risk management, but realising them depends on governance, high-quality data and human oversight. More research is needed on out-of-sample performance in stress periods, causal effects on institutional outcomes, emerging-market contexts and generative AI.
