📋 Tender Overview
The banking industry is stepping from the digital era to the artificial intelligence era. According to Dr. Dao Minh Tu, competitive advantage no longer lies in the scale of technology investment but depends on data quality, management capacity and customer trust. Download Dr. Dao Minh Tu, Vice President and General Secretary of the Vietnam Banking Association, speaking at the Conference - Photo: VGPData quality determines the power of AIAt the conference AI and data management in banking operations organized by the Vietnam Banking Association on August 5, in Hanoi, Dr. Dao Minh Tu, Vice President and General Secretary of the Vietnam Banking Association emphasized that AI is no longer just limited to chatbots or automating repetitive processes. Instead, this technology has directly participated in core areas of banking activities such as trade finance, cross-border payments and asset management. In addition, AI also creates value in daily operations through the ability to hyper-personalize customer experience, while promoting financial inclusion by exploiting behavioral data to expand risk assessment capabilities, helping many customers who do not meet traditional standards have the opportunity to access reasonable capital. The current digital banking race is no longer a race on algorithms or technology investment scale, but a race on data quality, comprehensive management capacity and digital trust from customers, Dr. Dao Minh Tu said. According to Dr. Dao Minh Tu, in recent years, Vietnam has continued to perfect a comprehensive, strict legal corridor and approach international standards through Data Law No. 60/2024/QH15, Personal Data Protection Law No. 91/2025/QH15 and Decree No. 356/2025/ND-CP... Artificial Intelligence Law has established the core principle that AI serves people, taking people as the center. and does not replace the authority and responsibility of people. The SBV is finalizing a draft circular regulating safety, risk management and implementation conditions for the application of artificial intelligence in the banking industry to concretize operating technical standards. The Vietnam Banking Association always agrees with the SBVs operating orientation: Innovation must go hand in hand with system safety and a good governance system, not to inhibit development, but so that we can confidently accelerate in an effective way. responsible and safe, Dr. Dao Minh Tu affirmed. The SBV is completing a unified management framework for AI. Speaking at the seminar, Mr. Hoang Minh Tien, Deputy Director of the Information Technology Department (SBV) said that the draft Circular regulating safety, risk management and implementation conditions for the application of artificial intelligence systems in the banking industry is built towards establishing a unified management framework, ensuring that AI is deployed safely, under control and in accordance with the specific needs of the industry. hostile to banking activities. Download Mr. Hoang Minh Tien, Deputy Director of the Department of Information Technology (SBV) - Photo: VGP According to Mr. Hoang Minh Tien, in fact AI has been widely applied in the industry, from credit scoring, loan appraisal, fraud detection, anti-money laundering (AML), customer care chatbot, eKYC to electronic authentication. However, along with the benefits it brings, AI also poses many risks in information security, model risk, data and responsibility in the decision-making process. This is also the practical basis for the State Bank to develop a draft Circular. Mr. Hoang Minh Tien said that the draft has been consulted many times and continues to be updated after Decision No. 33/2026/QD-TTg was issued. The draft includes 5 chapters, 26 articles, focusing on three major content groups: ensuring safety, managing risks and conditions for deploying AI systems. According to the draft, mandatory regulations will apply to AI systems that make decisions or support decision making in banking operations providing services to customers. For AI systems serving internal management and operations or being deployed within the framework of a controlled testing mechanism (sandbox), the draft aims to encourage application. One of the key contents of the draft is to classify AI systems into three levels including high risk, medium risk and low risk. According to Mr. Hoang Minh Tien, for high-risk AI, the draft closely follows Decision No. 33/2026/QD-TTg. In the banking sector, there are two groups of AI systems that are considered high risk: AI systems that automatically perform electronic transactions in banking activities and AI systems that automatically decide to grant credit. For payment transactions, AI systems are determined to be high risk when automatically performing transactions from 10 million VND to less than 100 million VND for individual customers or from 50 million VND to less than 500 million VND for institutional customers. According to Mr. Hoang Minh Tien, AI is only allowed to automatically make decisions within these thresholds. For transactions whose value exceeds the threshold, the system will not be allowed to automatically make decisions. Regarding safety requirements, Mr. Hoang Minh Tien said that the draft continues to inherit current regulations for information technology systems, while adding specific requirements for the system.h for AI. For cybersecurity, the draft requires testing and assessing specific risks of AI systems such as data poisoning, model poisoning, evasion attacks, model extraction attacks and data privacy invasion attacks. Regarding data, the draft stipulates that data used for AI must have a clear origin, ensure quality, be safely managed and control the risk of data leakage. The use of data must comply with the provisions of the Data Law and the law on personal data protection. Regarding risk management, the draft requires the development of an AI risk management framework in sync with the overall risk management policy and i
