Beyond AI Performance: A Unified Framework for Analyzing AI Integration in Financial Technology

Graduate School of Business HSE University Moscow, Russia This email address is being protected from spambots. You need JavaScript enabled to view it.
Graduate School of Business HSE University Moscow, Russia This email address is being protected from spambots. You need JavaScript enabled to view it.

Abstract

Artificial intelligence adoption in fintech is accelerating, yet successful deployment remains uneven in regulated and risk-sensitive environments. The relationship between AI architecture and sustainable organizational integration remains unclear. This study examines how fintech firms deploy classical machine learning, large language models, retrieval-augmented generation, and agentic systems using a combined Technology–Organization–Environment and Socio-Technical Systems framework. We analyzed twelve public cases across technological, organizational, environmental, and interaction dimensions. The presence of large language models did not imply high autonomy. Large-scale deployment was associated with systemic governance and calibrated trust, providing a structured basis for assessing AI deployment beyond model performance.