Page 494 - ISC PROCEEDINGS 21.4

Financial indicators—specifically CIR, CASA, and profitability—have shown significant
improvements attributable to AI implementation.
Data analysis and Processing methods
Once collected, the data is processed using the following techniques: (1) Content
Analysis: Extracting core themes regarding AI applications from management reports and
press releases. (2) Comparative Analysis: Utilizing tables to contrast the fundamental
differences between traditional banking models and AI-era banking, as well as comparing
the strengths and challenges of domestic banks. (3) Interdisciplinary Approach: Examining
the impact of AI not only from an economic standpoint but also regarding legal liability
and algorithmic ethics.
3. Results and discussion
3.1. Theoretical framework
Conceptual Definitions, drawing from the four conceptual frameworks discussed in
the preceding theoretical section, the author proposes a comprehensive definition of
banking within the digital era:
Banking in the AI era is defined as an intelligent financial services ecosystem that
operates on a foundation of Big Data and advanced algorithms. Its objective is to deliver
seamless, highly personalized, and maximally secure financial solutions, optimizing both
customer value and organizational performance through the synergy of human and
artificial intelligence.
In other words, banking in the AI era is a data-driven, autonomous system where
core banking functions—such as credit extension, risk management, and investment
consultancy—are primarily executed by machine learning and generative AI algorithms,
shifting away from a reliance on manual human intervention.
Consistent with the theoretical foundation discussed above, this study adopts the
framework focusing on the core contents and challenges of banking in the AI era.
3.2. AI application outcomes in selected Vietnamese banks and emerging
challenges
3.2.1. Outcomes of AI implementation in Vietnamese banking
Vietnamese banks are by no means outsiders in the global race. Over the past five
years, a robust transition from "Digital Banking" to "AI-First Banking" has emerged. This
section highlights practical AI applications among the pioneering banks in this field.
Electronic Know Your Customer (eKYC): This represents the most prevalent and
successful application, enabling banks such as TPBank, MB, VPBank, and Vietcombank to
achieve exponential growth in their user bases. These institutions utilize anti-spoofing
technologies, face matching, and Optical Character Recognition (OCR) to extract data
from citizen identification cards. By comparing real-time portraits with document photos,
AI ensures identity authenticity, thereby safeguarding customer security and financial
transactions. Implementing this AI technology allows for account opening within just five
minutes, eliminating the need for physical branch visits. Furthermore, advanced AI
models have been integrated to detect fraudulent behavior during the capturing process,
such as identifying coercion or 'screen-on-screen' spoofing attempts
Virtual Assistants and Intelligent Chatbots: Departing from the primitive scripted
chatbots of the past, banks have upgraded to AI systems powered by Natural Language
Processing (NLP). VietinBank, for instance, utilizes a virtual assistant that enables
customers to execute fund transfers and balance inquiries via voice commands directly
within the app. VIB leverages AI to personalize credit card payment reminders and

493