AI, skills and the future of work: A Bangladesh perspective

I can recall a time when I had to visit an export-driven factory where a routine that was very familiar to me suddenly became strange. Next to a newly installed AI-based inspection system on the production line stood a quality supervisor—a person who had spent years developing his judgement through experience.

Throughout most of his career, that judgment had been decisive: find flaws, accept batches, and guarantee quality before delivery. On this particular day, however, the screen beside him had the final word. It was able to scan thousands of garments in a few minutes, identify anomalies, anticipate the likelihood of rejection, and recommend corrective measures—much more quickly and reliably than a human being.

He did not complain when I asked him how his work had changed. He was not opposed to technology. He paused and grumbled, "Sir, I am still at work—but now I must know what the machine is telling me before I can take action."

No one had lost their job. No layoffs had occurred. However, the seat of power had changed. Manual judgment had been replaced by algorithmic insight. Experience could no longer be considered sufficient; learning had become mandatory.

This was one of the moments that defined the essence of artificial intelligence (AI) for Bangladesh. AI is not arriving as a job-killer. It is arriving as a skill-divider. People who can interpret systems, challenge outputs, and optimise processes will remain relevant. Individuals valued only for routine operation will quietly fall behind.

During my years of service—both in the military in uniform and later in government administration and corporate management—I have witnessed numerous technologies emerge and disappear. This has always been the case: machines progress, but it is people who decide whether they progress alongside them.

This reality has a direct impact on Bangladesh's workforce.

The discourse on AI in Bangladesh is still dominated by the fear of job loss. The question now echoing across factory floors, bank branches, call centres, and college corridors is whether machines will replace human work. Yet what is currently unfolding is a more specific and more consequential disruption.

AI is not eliminating jobs wholesale across sectors. Instead, it is gradually displacing those who are unprepared—those trained to repeat the same task endlessly and unable to evolve as job definitions change.

For a labour-rich country such as Bangladesh, this is not a hypothetical concern. It is a matter of competitiveness and belonging in the global economy.

Bangladesh's economy has been built on a manufacturing scale. Goods exports have surpassed USD 48 billion, largely fuelled by labour-intensive industries. This is an achievement worth celebrating. However, scale alone lacks depth.

At the same time, exports of ICT and digitally delivered services stand at around USD 1 billion—far behind regional competitors. This imbalance poses a strategic risk as global value chains become increasingly data-driven and AI-enabled.

Labour market pressure is already visible. Each year, large numbers of young people enter the workforce with academic qualifications but without job-relevant skills. Youth unemployment remains high. The rise of AI further widens this gap by reducing the value of routine, low-skill entry-level work and increasing demand for skilled, analytical, and adaptable roles.

It is important to be precise. Artificial intelligence does not destroy careers; it replaces tasks. The first activities to be automated are repetitive, predictable, and rule-based. In manufacturing, AI is used to optimise, predict, and inspect. In logistics, algorithms determine routing and inventory. In banking, AI analyses transactions, flags risks, and handles basic customer interactions.

People are still required—but only when they contribute judgment, supervision, and improvement in areas where machines fall short.

Regional comparisons are instructive. India's technology exports exceed USD 200 billion annually. Even amid economic strain, Pakistan has rapidly increased its ICT exports. Sri Lanka, despite having a much smaller population, earns more from digital services than Bangladesh. Vietnam has embedded digital capabilities into production, strengthening its position in global value chains. These outcomes reflect long-term investments in skills.

Policy options

First, skills must be treated as national infrastructure. Continuous reskilling is as essential to industry as ports and power. Digital literacy, data interpretation, and familiarity with AI tools should become standard across all sectors.

Second, industry must take responsibility. Companies that deploy AI should also retrain their employees. This should be supported by incentives that offset both the cost of skill development and the implementation of automation.

Third, education must move from being certificate-based to capability-based. The future demands flexible minds—individuals able to learn, unlearn, and relearn.

Fourth, the state must lead by example. To regulate, guide, and safeguard an AI-driven economy, public institutions need to develop strong digital capabilities.

Finally, employees must be supported through transitions rather than shielded from change. This is not about halting technological progress, but about preparing people before replacement becomes inevitable.

Bangladesh is not going to "get" AI. It is already here. It will not impoverish large segments of the population overnight, but it will quietly reward preparedness and penalise stagnation. The population can be transformed into a capable workforce only while the country's demographic advantage still holds.

The new reality of work is not a battle between humans and machines. It is a test of preparedness. Bangladesh still has time to meet that test—but that time will not last forever.

Sketch: TBS

Md Nazrul Islam is a former Executive Chairman of BEPZA, a retired Major General in the Bangladesh Army, and a PhD researcher into technology, workforce transformation and industrial competitiveness.

Disclaimer: The views and opinions expressed in this article are those of the author and do not necessarily reflect the opinions and views of The Business Standard.