Banking on AI: Bangladesh Prepares New Policy Framework
A national Artificial Intelligence (AI) policy is being drafted for the country’s banking and financial sector to strengthen fraud prevention, risk management, economic forecasting, internal operational efficiency, and customer service enhancement. The policy, to be issued by Bangladesh Bank, will provide guidelines on the use of Machine Learning (ML) and Large Language Models (LLMs).
Under this new framework, the central bank has already formed a seven-member team under its ICT Department to explore the deployment of a proprietary LLM system to minimize risks associated with cross-border data transfer. The committee is expected to prepare a draft policy within the next two months.
The committee is headed by Dr. Nurullah Shaheen, Joint Director (ICT) of the Enterprise Risk Management Department. Other members include Deputy Directors (ICT) Md. Belal Hossain and Md. Nurul Islam, Assistant Directors Mohammad Yasbir Chowdhury, Kazi Md. Ariful Haque, Md. Saeed Hasan, and Sayed Abrar Zawad. The committee has been assigned a defined operational scope and has already begun reviewing AI policy frameworks adopted by central banks in other countries.
According to internal sources, the policy will emphasize securing internal banking data to prevent any unauthorized exposure or misuse. It will ensure that AI-driven decision-making aligns with fairness, transparency, and ethical standards. Special focus will be placed on staff training to enhance AI usage skills across the financial sector.
The use of AI will extend to assessing borrower financial capacity, monitoring market, credit, and liquidity risks, and analyzing large-scale economic data for timely insights into growth trends, inflation, and foreign currency stability. AI-powered cyber monitoring systems will be introduced to prevent system intrusion, fraudulent transactions, and hacking attempts. In addition, AI will play a role in customer behavior analytics, fraud detection, and delivery of personalized services. Automated AI-based compliance monitoring systems will also be deployed to oversee policy and regulatory implementation across the banking ecosystem.
Regarding the initiative, Zakir Hasan, Executive Director in charge of Information and Communication Technology Department-2, told Amader Shomoy, “The objective of this policy is to ensure safe, appropriate, and regulated use of AI in the banking sector in the coming years. The policy will guide proper usage of AI both within the central bank and across all commercial banks. The approach is toward limited and well-governed implementation.” Responding to a question, he added, “This will not shrink employment; rather, it will expand opportunities. No one will lose their job. Those who gain AI skills will find employment more easily.”
According to data from the Bangladesh Institute of Bank Management (BIBM), 60 percent of banks currently do not have AI-integrated cybersecurity policies, while 40 percent do. Similarly, 68 percent have not embedded AI usage into their corporate governance frameworks.
Research findings indicate that 69 percent of banks are partially prepared for AI adoption, 11 percent are nearly prepared, and another 11 percent are fully prepared. In contrast, 9 percent of banks are not prepared at all. At present, only 5 percent of banks use AI in disaster recovery planning, while the remaining 95 percent do not apply AI in this area.
