HÀ NỘI – Artificial intelligence (AI) is set to become a key driver helping Vietnamese banks enhance efficiency, security, service personalisation and international integration, participants heard at the seminar ‘Building an AI-Powered Bank: A Revolution in the Banking Industry’ on Wednesday in Hà Nội.

A 2024 McKinsey survey found that 65 per cent of global organisations are already using AI regularly, particularly in data analytics and risk management. Deloitte also predicts that leading banks worldwide could boost business performance by 27–35 per cent thanks to AI, with average revenue per employee increasing by up to US$3.5 million by 2026.

Director of the Payment Department of the State Bank of Vietnam (SBV) Phạm Anh Tuấn said AI has become an indispensable tool in banking operations. Digital transformation in the banking sector has achieved remarkable results in recent years. Non-cash payments have grown rapidly, with the number of online and mobile transactions multiplying. Many commercial banks have proactively invested in digital platforms, developing chatbots, virtual assistants and AI-powered applications for credit assessment and fraud management.

When asked about the driving forces behind this shift, Dr Nguyễn Quốc Hùng, Vice Chairman and Secretary General of the Vietnam Banks Association, highlighted several factors: shrinking traditional profit margins; mounting operating costs requiring deeper restructuring; and growing customer expectations—particularly from digital-native generations—for predictive, instant and personalised services.

At the same time, competition from fintechs, bigtechs and e-commerce platforms is intensifying, while fraud, cyberattacks and money laundering are becoming increasingly sophisticated. Regulatory frameworks are also evolving, with stricter requirements for model transparency, accountability and risk governance. These dynamics are pushing the banking industry towards AI as a foundational pillar rather than a mere utility.

According to industry estimates, preparing data accounts for 15–25 per cent of a project’s budget, and up to 30–40 per cent in complex banking projects. Some reports suggest as much as 70 per cent of AI-related costs are hidden and unforeseen. Building and operating a centralised data hub can consume $3–4 million annually, with processing expenses making up 50–60 per cent of that figure.

Despite these high costs, Phan Thanh Đức, Senior Lecturer and Head of the Faculty of IT and Digital Economy at the Banking Academy, observed that while 90 per cent of enterprises invest in AI, only 1 per cent reach maturity in its application.

“How can we avoid the failure trap?” he asked, noting that commercial banks possess massive datasets from 20–30 years of credit history and customer behaviour. Yet having data and using it effectively with AI are two very different things. “Banks face heavy regulatory pressure and competition from fintechs, so major challenges remain in leveraging AI effectively,” Đức added.

From the banking side, a representative of ACB admitted that although the bank has collected vast amounts of data, the real challenge is how to exploit it effectively. The bank is currently working to integrate existing systems with new platforms, balancing ongoing investments with business innovation.

To help credit institutions overcome challenges and seize opportunities, the SBV’s Payment Department has identified four priority tasks for the coming period: completing regulations to encourage safe AI adoption; developing digital and data infrastructure; investing in financial and human resources; and strengthening co-operation with fintechs, bigtechs and international organisations. At the same time, financial literacy will be promoted to ensure citizens can use digital banking services safely and effectively. — VNS