Artificial intelligence (AI) is reshaping economies at speed, bringing both promise and peril, and Bangladesh must prepare early rather than react late, said Washington-based Bangladeshi analytics researcher Rokhshana Parveen.
“AI simultaneously creates efficiency and vulnerability. The countries that invest in governance, skills and data discipline now will define the next decade,” she said in an interview with TIMES of Bangladesh’s Mahfuz Ullah Babu.
Parveen, who spent more than two decades in Bangladesh’s commercial banking sector, is now applying her experience to advanced analytics research in the United States.
She previously held senior roles at Social Islami Bank PLC and Southeast Bank PLC, overseeing international trade finance, treasury operations, foreign exchange, remittance management, compliance and regulatory reporting.
Those roles gave her direct exposure to how fragmented databases and operational gaps create systemic risk.
“In banking, you learn quickly that delayed data is dangerous data,” she said.
After completing her Master of Business Administration (MBA) in Business Analytics in the United States, she is now pursuing a Doctor of Business Administration (DBA) in Business Analytics and Data-Driven Decision Management at the University of the Potomac in Washington DC.
Her doctoral research focuses on explainable artificial intelligence frameworks for public systems.
She said Bangladesh should prioritise three areas to stay competitive: national data infrastructure, ethical AI governance and diaspora engagement.
“First, we need a unified public data architecture. Ministries cannot operate in silos if we want predictive oversight,” she said.
“Second, AI regulation should not begin with restriction but with standards — auditability, fairness testing and human oversight must be mandatory.”
She also suggested establishing a structured global advisory network of non-resident Bangladeshi professionals.
“There is a growing number of Bangladeshis contributing to advanced research worldwide. A formal digital collaboration platform could allow policymakers to tap into that expertise in real time,” she said.
Her research targets fraud, waste and abuse in systems such as Social Security, Medicaid, Supplemental Nutrition Assistance Program (SNAP) and financial aid programmes in the United States. She is developing models that detect deepfake manipulation and behavioural biometrics patterns early, while reducing false accusations.
“One of the biggest risks of AI is opaque decision-making. If a system flags someone, it must explain why. Otherwise public trust collapses,” she said.
In a peer-reviewed paper published in the International Journal of Applied Mathematics in August 2025, she proposed a hybrid Explainable AI framework for ethical fraud prevention in United States federal welfare programmes.
Another study, published in October 2025, introduced a Behavioural Trust Vectorization Engine that achieved a 96.5 per cent recall rate in detecting identity theft by analysing user interaction patterns such as typing cadence.
Behavioural analytics shifts the focus from transaction value alone to interaction behaviour, she said. “Fraud is often behavioural before it becomes financial.”
Beyond fraud prevention, Parveen has researched sustainable finance, including the cointegration of green and non-green cryptocurrencies, linking financial innovation with environmental accountability.
She said Bangladesh must prepare its workforce for the AI transition.
“We should integrate data literacy and algorithmic ethics into university curricula. AI readiness is not only about engineers; it is about policymakers, auditors and regulators understanding the systems they supervise,” she said.
Drawing on 23 years of professional experience, she said cross-border fraud and compliance challenges now demand global coordination.
“Financial crime does not respect geography. Our regulatory thinking must be equally borderless,” she said.
Ethics remains central to her work.
“AI-driven decision systems in public infrastructure must be fair, transparent and auditable. Technology should strengthen institutional integrity, not weaken it,” she said.
Her long-term goal is to design secure, data-driven systems that government agencies and research institutions can deploy to protect public funds and reinforce trust.
“If we build AI with accountability from the start, it becomes a shield,” she said.
Alongside her present role in the United States, she said she wants to contribute to Bangladesh’s AI preparedness and policy development efforts.
