The scaling effect of digital risks in the banking sector: challenges of BigTech-concentration and artificial intelligence

Abstract

The purpose of this article is a comprehensive study of the theoretical and practical aspects of the emergence and deployment of the rapid scaling effect of digital risks in the modern banking sector, as well as the identification of systemic challenges caused by market concentration around BigTech providers and the rapid implementation of artificial intelligence technologies, to develop proactive tools for ensuring financial stability.

The study utilized a set of general scientific and special methods, including: methods of systemic-structural and comparative analysis – to study the nature of financial crises transformation; statistical-economic and trend analysis of the National Bank of Ukraine data – to assess the dynamics of cashless payments; tools of economic-mathematical modeling and index analysis – for the decomposition of integral indicators of systemic cyber vulnerability; the method of logical generalization – to form conclusions and recommendations for improving regulatory policy.

The non-linear nature of digital threats in the financial environment is investigated and the phenomenon of “bank sprint” is identified, which significantly reduces the time lag of depositors’reaction due to the use of mobile applications and coordination of panic moods in social networks. The dynamics of payment card transactions in Ukraine for 2023–2025 are analyzed, which confirmed the critical dependence of the national economy on the stability of digital communication channels. The structure of systemic cyber vulnerability is determined through the interaction of adversary, technological, and financial components in the SCyMoN model. The operational risk profile of key third-party ICT service providers (Microsoft, Google, Cisco, Apple, SAP, Salesforce) was constructed, proving a high level of co-movement between the vulnerabilities of providers and commercial banks. New threats of industrial use of generative artificial intelligence by cybercriminals are highlighted and the experience of implementing proactive data-driven risk management models, European regulatory standards, and the functioning of the domestic autonomous network POWER BANKING was evaluated.

It is proved that neutralizing the effect of rapid scaling of digital risks requires abandoning reactive approaches in favor of dynamic monitoring of operational resilience. The necessity of consolidating the efforts of regulators and banks in three strategic areas is substantiated: harmonization of national regulations with European legislation, deployment of proactive machine learning systems to detect anomalies, and expanding autonomous infrastructures of dual (physical and virtual) purpose to counter structural shocks.