Somewhere in Sylhet, a young woman is still waiting for the job she trained for. She spent three years studying data science, taught herself Python, built projects on her own, and graduated last spring carrying exactly the kind of ambition Bangladesh says it needs. Yet a year later, she is still searching for work. Not because she lacks talent or determination, but because the industry she prepared for has yet to fully emerge, while the one that is emerging was never fully reflected in her education.

Her story is far from unique. Nearly 885,000 graduates in Bangladesh remain unemployed, while graduate unemployment stands at 13.5 percent—higher than the national average. During the same period, the country’s IT export earnings grew by 13 percent. Employers say they cannot find the skills they need, while graduates struggle to find the opportunities they were promised. Somewhere in between lies a gap that should never have existed.

That gap—between what we keep announcing and what we actually build—is where our attention should be focused. And this year’s national budget, despite its good intentions, has not bridged it.

To be fair, the FY2026–27 budget did not ignore the sector. VAT exemptions for freelancers and content creators are practical and welcome measures. Skills development also received attention. None of these initiatives are insignificant.

But a few budget provisions do not amount to a national strategy.

A strategy requires institutions. It requires accountability. It requires someone within government whose explicit responsibility is to answer a fundamental question: where does Bangladesh want to stand in artificial intelligence by 2035, and what is the roadmap to get there?

Yet today, no one in government is explicitly tasked with answering that question. That absence—a structural gap in how we govern emerging technologies—is quietly costing us more than we acknowledge.

Consider the countries Bangladesh often looks to for inspiration. India did not simply allocate funding to its IndiaAI Mission; it created institutional structures, established research mandates, and tied implementation to measurable sectoral goals. The UAE went a step further by appointing a Minister of State for Artificial Intelligence. Singapore provided businesses, researchers, and citizens with a clear framework within which to innovate and operate.

These countries did not stop at recognizing AI’s importance. They built the machinery needed to act on that recognition. Bangladesh has made the declaration more than once. Building the machinery remains the unfinished task.

What would a serious response look like? At minimum, it would require four concrete steps.

First, Bangladesh needs a National AI Authority—a legally mandated body empowered to coordinate efforts across ministries, with representation extending beyond government officials. Its role would be to set priorities, oversee implementation, and publicly report on progress. Without a permanent institutional home, AI policy risks remaining a conversation rather than becoming a commitment.

Second, the country needs a National AI Strategy built around measurable outcomes rather than broad aspirations. Agriculture, healthcare, education, and public finance should each have clearly defined targets, designated lead ministries, and timelines for delivery. Farmers could benefit from AI-powered advisory services. Public hospitals could be supported by diagnostic tools. Tax authorities could deploy effective fraud detection systems. These are not futuristic concepts; they are solutions already being used elsewhere.

Third, Bangladesh needs a dedicated National AI Fund. Financing should be visible, transparent, and purpose-driven—not buried within broader ministry allocations. Such a fund could support university research, enable startups working on local challenges, and help public institutions adopt AI responsibly. Bangla-language models, agricultural intelligence platforms, and diagnostic systems designed for Bangladeshi communities will not emerge on their own. They require investment.

Finally, Bangladesh needs a Responsible AI Charter. Any system that influences decisions affecting citizens—whether in employment, credit, or public services—must be transparent, fair, and accountable. Responsible governance is not an obstacle to innovation; it is what builds public trust in innovation.

One concern inevitably arises in discussions about AI: the fear that it will destroy jobs.

History offers a more nuanced perspective. Every major technological transformation—from the printing press to electricity to the internet—disrupted existing forms of work while creating entirely new opportunities that few had anticipated. Technology itself was rarely the problem. The real challenge lay in whether institutions and workers were prepared to adapt.

Bangladesh’s young population is not a liability in the age of AI. Managed wisely, it could become one of the country’s greatest advantages. The graduate in Sylhet already possesses the determination and potential to succeed. What she lacks is a labour market—and a policy framework—that has prepared for her future.

The FY2026–27 budget has already been finalised, and it cannot be rewritten. What can still be shaped, however, is the year ahead. Establish the Authority. Develop the strategy. Launch the fund. Publish the charter. None of these actions requires waiting for another budget cycle. Much of this can begin through executive leadership, inter-ministerial coordination, and the political will to treat AI as the priority it has already become.

Bangladesh has never lacked ambition. What it has sometimes lacked is the follow-through that transforms announcements into institutions, and institutions into results.

The young woman in Sylhet is still waiting. So are thousands of others like her. Bangladesh has the talent to compete in the AI era; what it needs now is the leadership to prepare for it.

Reyad Hasnain is a policy analyst specializing in digital governance and public-sector reform.