The Ministry of Education has announced that an artificial intelligence (AI) curriculum will be introduced from Class III onwards from next year (the 2026-27 academic session). The aim is to prepare India’s future workforce for a “technology-driven economy”. Many state boards of education are currently working on AI curricula.
In July 2025, the government launched the SOAR initiative (Skilling for AI Readiness) through which now nearly 18,000 schools affiliated to the CBSE are already offering AI as a “skill subject” from Class VI onwards, with 15-hour modules for classes VI to VIII, and full-length 150-hour elective courses for classes IX to XII. The CBSE is said to be developing a framework for “AI integration across grades”.
This situation is in stark contrast with AI in higher education institutions. Only this year have a handful of universities introduced mandatory AI courses meant only for Science and Engineering students. It is not mandatory across disciplines.
One must, therefore, ask the question: Why this rush for AI in schools? It is indeed true that AI is set to make a great impact on our lives, and that the country seeks to “play a leading role in AI technologies and shape global AI standards”. Is teaching AI to primary- and middle-school children necessary or sufficient for building such leadership capability?
According to the government, the SOAR initiative helps to “bridge the digital divide” and creates opportunities for children “from rural areas or communities with limited resources”. In a country where an overwhelming majority of children and teachers have never used any digital tools in education, it seems at best ironical, and at worst sheer callousness, to talk of teaching AI as a means of bridging the divide.
To consider an analogy, the mobile phone has been the single most disruptive technology in the last three decades, and has already transformed society and the economy. Must we teach how mobile phones work to schoolchildren from Class III onwards? Schools must certainly engage with mobile-phone usage by children, but at an age when they are able to comprehend it better. They must teach the children to use it safely, responsibly and for its educational value.
Often, the public use of the phrase “AI in schools” hides considerable confusion over what is meant by it. Some mean a vaguely defined form of AI literacy. Some want increased use of AI tools in classrooms. Others mean the use of AI to enhance “teacher productivity” (for example, to prepare class presentations). AI developers speak of personalised learning and assessment. Governments talk of using AI to track every child’s academic progress individually. In such a confused situation, it is imperative that we separate informed use of AI tools from teaching AI to children.
Coming to the specifics of the currently implemented curriculum, the middle-school AI curriculum introduces three AI domains: Computer vision, natural language processing (NLP), and statistical data. The Class VII text also speaks of the “innovative role of AI in fostering sustainability and societal development, highlighting key concepts like Sustainable Development Goals, system thinking, and system maps”. In Class VIII, children learn “the AI Project Cycle”, AI ethics and responsible AI practices. Class IX discusses mathematics for AI and generative AI. Class X includes topics such as supervised, unsupervised and reinforcement learning models, clustering and neural networks.
Looking at this collection of topics, it is hard to understand how children would put all this together in their heads in relation to the AI apps they would use in which these learning models are deeply embedded and realised in software. How would they relate all this to the Mathematics and Science they are learning? How does a 12-year old make sense of “fostering sustainability and societal development” or for that matter the use of system maps? These are notions that require considerable maturity. What kind of pedagogy is needed to teach such concepts to children?
Set against such considerations, consider this question from the Class VII AI curriculum handbook of CBSE:
Which SDG focuses on ‘gender equality and empowering all women and girls’?
(a) SDG-3 (b) SDG-5 (c) SDG-8 (d) SDG-10.
Does it matter which number it is? If learning in school about how AI works is to help children develop a critical outlook on the use of AI, exercises of this kind nullify that hope. Given how little success we have had so far on critical thinking in our Science education, such expectations seem unrealistic.
The question of teaching AI in schools is not to be posed as one with a yes or no answer. It is about examining the educational purpose, the pedagogic means and the assessment modes for doing so, and developing teacher capacity and educational resources for achieving the intended purpose. In the case of children in primary or middle schools, the psychology of children’s learning is also critical.
In the specific case of AI, there are other considerations: Is it a mere technological bubble? This is a point of debate among computer scientists. School is a learning space for building strong foundations. Being pressured by technological booms may not serve the educational goals of society in the long run.
As a technology, AI is both seductive and addictive. We need the exercise of social wisdom, rather than clamour, in taking decisions on AI in the educational space and while placing AI in the hands of small children. Above all, we must keep asking ourselves: Are we being wise?
The writer is professor, Azim Premji University, Bengaluru, and faculty (retired), Institute of Mathematical Sciences, Chennai
