Naima Onamika
Commissioning Editor, TOB
India’s Head Start in Policy Preparedness
India’s greatest advantage is that it had already established a policy foundation for modernising its education system before the arrival of generative AI. The National Education Policy 2020 placed strong emphasis on critical thinking, problem solving, and practical skills. In addition, the Central Board of Secondary Education has taken initiatives to introduce AI related education at the school level. Various national initiatives have also been introduced to strengthen AI capabilities.
However, having a policy framework on paper does not automatically ensure the adoption of AI in practice. While leading private schools and technology driven institutions in major Indian cities are relatively advanced in their use of AI, many government and rural schools still lack reliable internet access, adequate devices, and trained teachers. India therefore faces the risk of creating a new form of digital inequality in AI enabled education.
Bangladesh’s Challenge: Rote Learning and AI
The challenge is more severe in Bangladesh. According to various sources and analyses published by experts, the country’s education system remains significantly dependent on rote learning.
In such a system, student assessment often focuses more on final answers or written results than on their ability to think, investigate, and demonstrate reasoning. As a result, generative AI can easily produce an entire assignment or writing task on behalf of a student. The problem is not simply cheating through AI. The deeper concern is whether students are bypassing the process of thinking for themselves. For Bangladesh, therefore, the priority of AI policy should not be to ban AI, but to establish an education system that evaluates students’ thinking, research, analytical abilities, and creativity.
Teachers: The Most Important Weakness in Both Countries
The most important investment required for the successful integration of AI into education is investment in teachers. In Bangladesh, many teachers already work under the pressure of heavy workloads, limited salaries, and private tutoring. As a result, the time and institutional incentives required to learn new technologies are often lacking.
India has a relatively broader teacher training system, but it faces similar challenges. Administrative responsibilities and limitations in professional capacity within government educational institutions can hinder the adoption of new AI based teaching methods.
Teachers should not be trained only to use AI tools. They also need to understand when AI should be used, when it should not be used, and how AI generated information should be verified.
Fundamental Changes Are Needed in Assessment
Generative AI has exposed one of the education system’s greatest weaknesses in assessment. If students are simply asked to submit a 1,000-word assignment, AI can produce it within seconds. A different approach is possible when assessment includes:
• Initial research ideas
• Drafting and revision
• Personal reflection
• Oral presentation
• Viva or oral defence
• Classroom discussion
Under such an assessment model, AI cannot complete the entire learning process on behalf of the student.
Both India and Bangladesh therefore need to move from product-based assessment towards process-based assessment.
Clear Guidelines for Use of AI in Universities
The issue becomes more complex in higher education. Without clear policies defining permitted and prohibited uses of AI, students may face different consequences for similar practices across institutions.
Bangladeshi universities need to develop discipline specific AI policies. These should address questions such as:
• To what extent is AI assisted literature searching acceptable in research?
• What are the rules for using AI generated code in programming?
• To what extent should AI assisted translation or language correction be permitted?
• How should the use of AI in a thesis or dissertation be disclosed?
• Who bears responsibility for AI generated content?
India also needs institution and discipline specific implementation alongside national level guidelines.
Not Prohibition, but Responsible Adoption
The most important policy shift regarding generative AI in education should be a move from prohibition towards responsible stewardship. If AI is treated solely as a tool for cheating, students are likely to use it secretly. If AI literacy becomes part of education, students can learn what AI can do, what it cannot do, and when human thinking remains essential.
AI literacy curricula can be introduced for both teachers and students. At the same time, issues such as privacy, copyright, misinformation, and academic integrity should be incorporated as essential components of AI education.
Shared Policy Priorities for India and Bangladesh
The realities of both countries point to several common policy priorities:
• AI literacy should become part of the national curriculum.
• Regular and paid professional development programmes on AI should be introduced for teachers.
• Assessment systems should place greater emphasis on oral examinations, drafts, projects, and classroom-based evaluation.
• Universities should develop discipline specific policies on AI use.
• The digital infrastructure gap between rural and government educational institutions and their better resourced counterparts should be reduced.
• Governments, universities, and technology companies should work together to develop AI ready curricula.
• Alongside teaching students how to use AI, education systems must also develop their ability to identify and verify AI generated misinformation.
Conclusion
The arrival of generative AI in the education systems of India and Bangladesh is not simply a technological development. It has brought fundamental questions about how a student’s skills should be assessed to the centre of the education debate. India’s relatively stronger policy framework gives the country a degree of advantage, but significant challenges remain in implementation, including digital inequality, teacher training, and institutional capacity. Bangladesh faces a deeper challenge. Its dependence on rote learning and limitations in teacher preparedness increase the risk of turning AI from a supportive educational technology into a shortcut to learning.
The central lesson for both countries is the same. AI cannot be kept outside the education system, but education systems cannot be allowed to become dependent on AI either. What is needed is a policy framework in which AI does not replace students’ thinking but serves as a tool to make that thinking deeper, more creative, and more effective.
