AI Is Here. Can Bangladesh Keep Up?

Artificial intelligence (AI) is already transforming workplaces around the world. Through discussions with policymakers, businesses and technology experts in recent months, one message has been clear: AI is no longer a future possibility. It is already writing reports, translating languages, analysing data, generating software and handling customer queries. For Bangladesh, the challenge is no longer whether AI will change the way we work, but whether our workforce can adapt quickly enough.
This transition comes at a difficult time. Bangladesh is already struggling to create enough quality jobs, particularly for educated youth and women. In 2025, World Bank reported that half of Bangladeshis aged 15-29 were in low-paying jobs, while one in four educated young women remained unemployed, highlighting a persistent skills mismatch. AI could deepen these challenges unless workers are equipped with the skills needed for a rapidly evolving economy.
The World Economic Forum’s Future of Jobs Report 2025 estimates that global labour-market changes could create 170 million jobs by 2030 while displacing 92 million. More importantly, employers expect 39 per cent of workers’ existing core skills to change. AI and big data are among the fastest-growing skills, alongside human capabilities such as creative thinking, adaptability and collaboration.
South Asia is already seeing the shift. World Bank analysis of job postings found that the share requiring AI-related skills more than doubled between January 2023 and March 2025, from 2.9 to 6.5 per cent. Such jobs also commanded a wage premium of nearly 30 per cent over other professional positions. For Bangladesh, however, there is an uncomfortable detail. World Bank data show that Bangladesh accounted for only a tiny share of AI-related job postings captured in its South Asian dataset in 2025, well behind India and Sri Lanka. This should be read as an early warning.
The risk is often misunderstood. AI will not simply eliminate occupations. It will change tasks within occupations. An accountant using AI may become more productive than one who does not. A garment supervisor could use predictive tools to reduce waste. Teachers could prepare learning materials faster, doctors could receive diagnostic support, farmers could use better weather and market information, and government officials could analyse data more efficiently. The emerging divide may therefore be less between humans and machines than between workers who can use AI and those who cannot.
Bangladesh is not starting from zero. The World Bank-supported Accelerating and Strengthening Skills for Economic Transformation (ASSET) project, worth $300 million, was designed to equip more than one million youth and workers with future-oriented skills. The ADB-supported Skills for Industry Competitiveness and Innovation Program (SICIP), backed by $300 million from ADB and $75 million from the government, aims to reach around 220,000 workers between 2024 and 2029. 
These are major investments. But are our skills programmes changing as quickly as the skills themselves? Training systems often remain organised around courses, certificates and enrolment targets. AI demands faster curriculum updates, retraining of trainers, stronger employer involvement and greater attention to employment and wage outcomes. SICIP has already moved in this direction through advanced technology courses and industry partnerships. That approach should spread across the wider skills system.
Our neighbours are moving quickly. India’s IndiaAI Mission has an approved outlay of Rs 10,371.92 crore, including Rs 882.94 crore for IndiaAI FutureSkills. Pakistan’s National AI Policy, approved in 2025, includes plans for large-scale AI education and upskilling. Sri Lanka’s AI strategy similarly places workforce capability and AI literacy within its economic transformation agenda. 
Bangladesh is also stepping onto the international AI stage. Alongside developing its National AI Policy 2026-2030, the government has decided to join the newly established World Artificial Intelligence Cooperation Organization (WAICO) initially as an observer. Established in Shanghai in July with 29 founding countries, WAICO seeks greater international cooperation on AI, including narrowing the digital divide and supporting more inclusive AI development. For Bangladesh, observer status should be more than a diplomatic presence. It could open opportunities for knowledge exchange, capacity building, research partnerships and access to emerging technologies. International engagement, however, will matter only if it builds capability at home.
A practical step would be an AI and Future Skills Fund, initially around Tk500 crore annually. This is not an existing government commitment, but a policy benchmark worth considering. It could support AI literacy across universities, polytechnics and TVET institutions, retrain instructors, finance industry-led apprenticeships, develop Bangla-language AI tools and help small businesses adopt affordable AI applications.
Importantly, Bangladesh does not need to train millions of AI engineers. It needs millions of AI-capable workers. That distinction matters. A nurse does not need to build a machine-learning model, but should understand how AI tools can support care and where human judgement remains essential. The same applies to teachers, accountants, administrators, factory supervisors and entrepreneurs. Development partners can accelerate this shift. Future programmes should increasingly support employer-designed micro-credentials, AI-enabled apprenticeships, trainer development and regular labour-market intelligence. Inclusion must also be deliberate. Women, rural youth, persons with disabilities, migrant workers and those vulnerable to automation need targeted access to reskilling. Otherwise, AI could widen existing labour-market inequalities.
Bangladesh cannot outspend the world’s technology powers on artificial intelligence. It does not need to. Its strongest economic asset remains its people. But a demographic dividend is never automatic. In an AI economy, its value will increasingly depend on what those people know, how quickly they can learn, and whether our institutions can keep pace. AI is moving quickly. Bangladesh’s skills system must move quicker.
_____________________________________
The writer is an academician and development strategist, currently serving at a foreign diplomatic mission in Dhaka. He can be reached at [email protected]