Discuss the ethical and pedagogical challenges of introducing generative AI tools to primary school students in India.

India will make Artificial Intelligence and Computational Thinking (AI-CT) a compulsory subject from Class 3 in the 2026-27 session, anchored in NEP 2020 and NCF-SE 2023 [1][2][3]. While this builds AI-readiness, exposing children in the foundational stage to generative AI raises ethical and pedagogical concerns that must shape the rollout.

Ethical challenges

• Children's data privacy: Gen AI platforms profile minors. The DPDP Act, 2023 mandates verifiable parental consent for processing children's data, but educational institutions enjoy partial exemptions and no India-specific EdTech-AI compliance code yet exists [4].
• Age-appropriateness and human agency: UNESCO's Guidance for Generative AI in Education and Research urges an age limit for independent interaction with Gen AI and warns against erosion of human agency [5].
• Bias and hallucination: young learners lack the critical evaluation skills to detect fabricated or culturally skewed outputs, risking uncritical absorption of error.
• Equity: UDISE+ 2023-24 shows only 57.2% of schools have functional computers and 53.9% internet access [6]; unequal exposure can widen, not bridge, the digital divide.

Pedagogical challenges

• Cognitive substitution: the foundational stage builds literacy, numeracy and independent problem-solving; premature reliance on generative answers can weaken effort-based learning — the concern behind restrictive moves in some foreign school systems.
• Teacher capacity: simultaneous scale-up across CBSE, KVS, NVS and State boards demands retraining lakhs of teachers through NISHTHA, with uneven readiness [1].
• Tool over objective: AI aids must serve a defined learning outcome rather than substitute pedagogical judgment.
• Federal variance: school education being a Concurrent subject, State adoption of the central framework will differ, creating quality asymmetry.

The challenge is therefore one of sequencing, not prohibition. A graded model — computational thinking and AI literacy at the primary stage, supervised generative-tool use later, backed by child-specific data safeguards, teacher training and an outcome-based review mechanism — can deliver NEP 2020's vision of technology-enabled yet child-centred education.