Is AI Equaliser or Disruptor for Bangladesh’s Economy?

Artificial intelligence (AI) is rapidly becoming part of everyday life, transforming businesses, work and economies. At its core, AI refers to computer systems designed to perform tasks that traditionally require human intelligence. Unlike conventional programmes that follow fixed instructions, modern AI can analyse data, recognise patterns, learn from experience and support increasingly complex decisions. Yet its progress is outpacing the security measures and cyber policies meant to govern it, and the risks of misuse, from cyberattacks to large-scale disinformation and now autonomous “agentic” systems, are growing alongside its benefits. Countries that adopted automation and digital technologies early are already reaping the rewards: more efficient production and new industries. The central question for Bangladesh is not whether to adopt AI, but how to harness its economic potential without leaving millions of workers behind.

Historically, cutting-edge technology has favoured large corporations with deep pockets. AI may change that. By lowering the cost of accessing specialised knowledge and digital capabilities, it can become an important equaliser for Bangladesh’s small and medium enterprises (SMEs). A leather goods manufacturer in Hazaribagh or an agro-processor in Bogura cannot afford in-house compliance teams, data analysts or international market researchers. Generative AI can make some of these capabilities more accessible, helping smaller firms improve productivity, meet export standards and compete more effectively. AI also offers major opportunities for the software and IT-enabled services sector, where Bangladesh has a large pool of developers, freelancers and outsourcing professionals serving global clients. Rather than competing mainly on low-cost labour, firms could move towards higher-value services such as AI-assisted software development, data analytics, cybersecurity and customised digital solutions. The financial sector shows similar promise. AI-powered credit assessment can analyse transaction patterns and digital payment histories, potentially expanding access to finance for people without conventional collateral or credit histories while helping lenders manage risk.

AI also carries serious risks, particularly where Bangladesh is most vulnerable. The ready-made garment sector employs more than four million workers, most of them women. As global brands adopt automated design, robotic cutting and AI-driven inventory management, Bangladesh’s low-cost labour advantage may weaken. Without effective reskilling, many lower-skilled workers could face displacement. The challenge is compounded by a skills gap. More than two million young people enter the job market each year, yet what students learn often differs from what employers need. Unless education develops critical thinking, data literacy, ethical judgement and AI skills, many graduates may be poorly prepared.

Accountability is another concern. As algorithms increasingly influence decisions on loans, recruitment and healthcare, who is responsible when an AI system makes a costly or discriminatory mistake? A newer and more serious risk is agentic AI: systems that do not merely answer questions but plan and act on their own, writing code, accessing databases, sending messages and initiating transactions with little human supervision. The same autonomy that makes them useful makes them potentially destructive. An agent with broad access can misread an instruction, act on false information or be manipulated by malicious input, and a single error can spread across connected systems faster than any human can intervene. Attackers can also use them to scan for weaknesses, craft phishing messages and launch cyberattacks at unprecedented scale. For a country embracing digital banking, mobile financial services and e-governance, the consequences could include financial loss, disrupted services and eroded trust. Accountability is harder still when no human approved the specific action that caused the harm.

Indeed, technological revolutions do not simply destroy jobs; they transform work and create new ones. So, Bangladesh’s challenge is not to compete with AI, but to prepare workers to use it effectively, moving from routine tasks towards technical skills, critical oversight, data interpretation and creative problem-solving. This requires deliberate policy action. A national reskilling programme should be developed jointly by industry and technical training institutions, preparing vulnerable workers for new roles before, not after, they lose their jobs. Universities and polytechnics must also reform curricula to build data literacy and AI skills across disciplines.

AI can also help diversify the economy, from semiconductor design to renewable-energy forecasting and grid management. South Korea, Malaysia, Singapore and India show that diversification demands long-term investment in skills, technology and industrial capabilities. Bangladesh also needs clear rules for AI in areas where automated decisions have serious consequences, with defined accountability and mechanisms for review and redress. These rules should explicitly cover agentic systems, with limits on their permissions, human approval for high-stakes actions such as large payments or changes to critical records, and audit trails so every automated action can be traced and reversed. Smaller businesses should receive support in adopting AI so its benefits do not remain concentrated among better-resourced firms.

AI is neither a guaranteed solution to Bangladesh’s economic challenges nor an inevitable threat to jobs. Whether it drives inclusive growth and higher productivity, or widens inequalities, will depend on how the country prepares for and manages it. Universities have a key role. They should offer flexible short courses and micro-credentials so that students and professionals can acquire AI, data and emerging-technology skills. As AI moves from assisting people to acting for them, humans must remain firmly in control of the decisions that matter most. 

_____________________________________

The writer is a Distinguished Professor of Eastern University and former Vice-Chancellor of East West University