AI is already reshaping education worldwide. But in Bangladesh, the bigger question is not simply how we use AI in classrooms. It is: Which framework will guide us? Will we fall back into an ego-system -- defined by silos, competition, and distrust -- or can we build an eco-system of trust, alignment, and shared purpose that benefits all learners?
This piece is written with respect for the many champions across Bangladesh’s education system -- teachers, officials, innovators, and civil society actors -- who are working hard to improve learning. Yet we must be honest about the reality: Too often our system reflects an ego-system, where good ideas lose momentum amid fragmentation and politics.
The real challenge, then, is not whether Bangladesh will adopt AI in education, but whether we will choose the right frameworks to ensure AI strengthens, rather than weakens, our educational futures.
Frameworks to guide us
Globally, frameworks already exist that can help Bangladesh think through the responsible use of AI in education. UNESCO’s Readiness Assessment Model (RAM) looks at a country’s preparedness across five dimensions: legal, socio-cultural, scientific/educational, economic, and technical. Within education, RAM highlights strategy, infrastructure, curriculum, attainment, and access to AI learning opportunities.
Bangladesh has also taken steps in this direction. In 2041, education foresight exercises with government and non-government stakeholders, eight futures were identified: Learners, curriculum and pedagogy, policy, resources, assessment, teachers, governance and partnerships, and research and innovation. Similarly, the draft Blended Education and Skills Masterplan (BEFA) set out dimensions like teaching practices, resources, assessments, capacity, employment, and inclusive infrastructure, supported by enablers such as policy, financing, and partnerships.
It’s reasonable to ask: Should there be an AI in the Education and Skills Framework that mirrors these efforts? Or perhaps a broader Technology in Education and Skills Framework, so that future tools like quantum computing or extended reality are approached consistently, without reinventing the wheel each time?
Holistic dimensions
Frameworks are helpful, but they can also be limiting. The STEEPLE model (social, technological, economic, environmental, political, legal, ethical) reminds us that AI is not just a technical matter. Bhutan’s Green Schools framework (developed by the previous Minister of Education Thakur S Powdyel) is even more holistic, including dimensions like aesthetic, spiritual, and moral development.
What if we applied that same lens to AI? Beyond efficiency and innovation, should we also ask: What are the spiritual, aesthetic, and moral dimensions of AI in education? These questions may sound unusual, but they remind us that education is not only about skills -- it is about values, meaning, and human flourishing.
Foresight tools like Causal Layered Analysis (CLA), developed by futurist Sohail Inayatullah, take this further. CLA asks us to move from the surface level of news headlines (the litany), into deeper layers: Systemic structures, worldviews, and myths. In the case of AI, the headlines might speak of automation, disruption, or opportunity. The systems level looks at policies and governance. The worldview level asks whether we see AI as a threat, a partner, or a tool. And at the deepest layer, myths and metaphors shape our imagination -- do we view AI as a teacher, a rival, or perhaps even a mirror of ourselves?
By reflecting on these layers, we can avoid adopting frameworks blindly. Instead, we can recognize how they carry implicit worldviews and narratives -- and decide whether those serve Bangladesh’s context.
Mapping AI to education functions
One practical way to think about AI in education is to map AI functions against education system functions. For example:
- Perception (AI) - data collection (education)
- Reasoning - data analysis and insights
- Decision-making - policy and planning
- Communication - stakeholder engagement
- Creativity - content generation
- Action - implementation
This mapping shows that AI can be more than a back-end tool -- it can support educators and policymakers in every layer of the system, from classroom practice to national strategy. But the key is alignment. Without a shared framework, AI risks becoming another fragmented initiative that deepens the ego-system.
Cultural appropriateness
The next question is: What kind of framework is appropriate for Bangladesh? Three principles are worth keeping in mind.
- Alignment with existing frameworks: A new AI framework should not sit in isolation; it should connect with familiar approaches like BEFA or use language that is familiar to local education system actors
- Co-creation: Frameworks must not be decided by a few experts in isolation. They need to be developed collaboratively with teachers, policymakers, learners, and communities
- Cultural grounding: Any framework should incorporate local and indigenous values. What “local context” means will vary, but clarity here is crucial
The road ahead
Bangladesh has no shortage of vision documents, strategies, and masterplans. But the challenge is not only what frameworks we use -- it is how we bring them to life. That requires shifting from an ego-system of silos to an eco-system of trust and collaboration.
The first step is to ask the right questions. AI will shape education in Bangladesh whether we prepare for it or not. The choice is whether we allow it to deepen divisions, or use it to build bridges -- between agencies, between actors, and most importantly, between the present and the future our learners deserve.
If AI can help us dream together, then perhaps it can help us work together too.
Shakil Ahmed is an educator, futurist and storyteller at Ridiculous Futures, network coordinator at #NextGenEdu and studying his PhD in Futures Studies at Tamkang University, Taiwan.