Integrating AI into national curriculum calls for policy changes

Just recently, a young visually impaired SSC candidate from Thakurgaon expressed his sorrow on Facebook while looking for a scribe. Scribes for the visually impaired act as human writers during examinations, recording answers dictated by the student. In his post, Sharif Ali shared that without a scribe, his participation in the examinations would become uncertain.

His struggle raised widespread concern, but it also prompted a deeper question: in an age where technology can read, write, and respond in real time, why are our systems still so rigid? The answer lies not in a lack of tools, but in how slowly our education policies adapt to them.

Globally, artificial intelligence is no longer a distant concept in classrooms. It is already shaping how students learn, access information, and even think.

In Bangladesh, however, its presence has largely been confined to universities, often guided by rudimentary rules, with little evidence of meaningful, system-wide integration. However, earlier this year, Education and Primary and Mass Education Minister Ehsanul Haque Milon reportedly signalled a forward-looking shift, promising reforms that include introducing fields such as nanotechnology, AI, and robotics into educational programmes, alongside expanding technology-driven learning and digital platforms. If implemented with clarity and commitment, this vision could align Bangladesh’s classrooms with a rapidly evolving world.

One persistent gap in our current approach is the tendency to equate AI education with coding. This is a narrow interpretation of a tool that is inherently cross-disciplinary. Imagine history classes where students analyse migration patterns, or language classes where AI supports personalised writing feedback. These are not distant possibilities. In a system long shaped by a one-size-fits-all model, AI offers a pathway toward more competency-driven and adaptive learning.

However, this transformation hinges on a single, often overlooked factor: teachers. Many educators still have limited exposure to AI tools, let alone the frameworks needed to integrate them meaningfully. Some feel intimidated; others remain understandably cautious. Without addressing this, AI risks becoming another theoretical layer rather than a practical classroom tool.

Policy, therefore, must begin with teachers, not as passive recipients of reform, but as its primary drivers. A national teacher upskilling roadmap is essential, one that focuses on both technical familiarity and practical classroom application. AI modules should be embedded within teacher training colleges, while continuous professional development programmes can adopt blended learning models to ensure accessibility across both urban and rural contexts. At the same time, these efforts must be supported by targeted infrastructure improvements, particularly in connectivity and access to devices.

Equipping schools with low-cost laptops or tablets can create the foundation for digital learning, while community-based solutions, such as mobile learning labs that bring technology directly to underserved areas, can help bridge persistent gaps. Continued expansion of reliable internet access will further enable the effective use of AI tools, ensuring that these initiatives translate into meaningful learning outcomes rather than remaining limited in scope.

There are also compelling equity considerations. Take the case of Sharif Alam. According to his school’s headmaster, the institution has not been able to introduce Braille-based teaching due to the absence of specialised teachers, who are supposed to be provided by the Department of Social Services, despite repeated requests. While this gap is a separate issue, AI, if thoughtfully deployed, could help bridge such challenges through assistive technologies, converting audio-visual materials into accessible formats, or providing patient, responsive learning support that adapts to individual needs.

Structured collaboration between universities, technology firms, and innovation hubs will also be critical. Institutions like the Universal College Bangladesh (UCBD), already working with technology-sector partners, offer a glimpse of how such ecosystems can function, bringing real-world datasets, industry insights, and research opportunities into the learning process. These partnerships can also create pathways for internships and apprenticeships, ensuring that AI education does not remain confined to textbooks.

Finally, there must be mechanisms for continuous measurement and adaptation. Any meaningful integration of AI into the curriculum must also be grounded in ethics. Students and teachers alike need to be equipped not just to use these tools, but to question them, understanding issues such as data privacy and consent, recognising how algorithmic bias can shape outcomes, and identifying the risks of misinformation and deepfakes.

The promise of AI in Bangladesh’s education system is not abstract, it is tangible and within reach. But its success will depend on whether we centre our efforts on people, particularly teachers, rather than technology alone.

* Professor Muhammad Ismail Hossain is the Dean of Academic Affairs, Universal College Bangladesh.