AI Becomes Mandatory From Class 3 in 2026-27: A Practical Brief for Schools and Teachers

India's school system has been talking about AI for two years. From the 2026-27 session, the talk turns into a timetable line.

The Ministry of Education has confirmed that an integrated curriculum on Artificial Intelligence and Computational Thinking will be introduced for all schools from Class 3 onwards in the 2026-27 academic session. Classes 3 to 8 are inside the first wave; Classes 9 and 10 follow in 2027-28. The expert review for NCERT's draft curriculum is being led by Professor Karthik Raman of IIT Madras (policy explainer). Learning materials, teacher handbooks and digital resources were targeted for completion by December 2025.

What this changes on the timetable

Until now, AI sat in two places: as an optional skill subject from Class 9 in CBSE schools, and as a 15-hour SOAR (Skilling for AI Readiness) module offered by around 18,000 CBSE schools to Classes 6 to 8. The 2026-27 reform absorbs SOAR into a mandatory backbone, drops the entry point to Class 3, and stretches the strand across multiple subjects rather than treating it as a standalone elective (CBSE rollout reference).

The grade design splits responsibility. In Classes 3 to 5, Mathematics and other subject teachers handle the computational thinking strand. In Classes 6 to 8, teachers from multiple disciplines collaborate to integrate the strand across science, social science and language. Computer Science teachers will lead AI-focused instruction once Class 9 enters the wave.

Teacher training: NISHTHA, not a special course

Training will be delivered through the existing NISHTHA platform — the National Initiative for School Heads' and Teachers' Holistic Advancement — using grade-specific, video-based, time-bound modules. Schools should expect their teachers to be assigned NISHTHA enrolments through the state DIET pipeline, rather than receive a separate AI certification on the side.

Two practical implications. Every Maths teacher in Classes 3 to 5 needs to budget for the module before the academic year starts. And schools that have already invested in private AI training packages should now check whether those packages will be recognised as supplementary, or whether they will run alongside the official module.

What admins should plan for in May and June

The summer break is the only window left to put structure around the rollout. Five questions deserve a slot in the planning meetings.

The first is timetable architecture. AI in Classes 3 to 5 will not be a standalone subject the way Hindi or Maths is — it is a strand woven into existing periods. The academic head has to decide whether Maths absorbs most of it, or whether it is co-taught with EVS and the Computer period.

The second is teacher load. If the NISHTHA modules add 10 to 15 hours of preparation per grade across the year, that is real time. Schools that already run heavy weekend workshops will struggle to layer this in without trimming somewhere else.

The third is equipment. The early-grade computational thinking strand is not laptop-dependent. Logical reasoning, sorting, pattern recognition and simple programming logic can be delivered without devices. Schools without a one-to-one device programme do not need to panic — they need to be deliberate about the unplugged activities.

The fourth is parent communication. Parents of Class 3 to 5 students are about to see "AI" appear on the term plan. Most will benefit from a one-page note explaining that this is not coding boot camp, that there is no separate fee, and that the strand will be assessed alongside Maths.

The fifth is assessment. The Ministry has not yet published assessment guidance for Classes 3 to 5. Schools should treat this as a placeholder for now and avoid building large summative weights into the first year.

What teachers should do this term

For Maths and subject teachers in Classes 3 to 8, the next ten weeks are the highest-leverage window. Three concrete steps.

Bookmark the NISHTHA module ID for your grade as soon as it is released. Modules will be released in waves through state-level coordinators; the 2025-26 NISHTHA secondary cycle gives an indication of typical structure.

Build a small lesson sandbox. Pick one chapter you teach and try drafting a thirty-minute computational thinking sub-task that fits inside it. A Class 4 Maths chapter on patterns can host a sorting algorithm activity using physical cards. The point is to internalise the integration, not write a textbook.

Flag any device dependencies up to the academic head early. If a draft module assumes a tablet rotation that the school does not have, raise it before timetabling locks down. Implementation is grade-specific and resources need to follow.

What this means for parents

The 2026-27 timetable will look slightly different from the 2025-26 one. There will be no separate AI period in primary classes. There will be no separate fee. Children will not need to buy a device or a special textbook. Most of what changes will be inside Maths and EVS lessons, gently.

The mistake to avoid is over-investing in private AI courses for primary-school children. The whole point of moving the entry point to Class 3 is to make AI literacy universal and free. The premium-tutoring market will respond by selling worry; parents do not need to buy it.

The bigger picture

The 2026-27 cycle is the first time NEP 2020 commitments meet the regular school timetable at scale. NCERT's new Class 9 textbooks arrive the same year, the National Curriculum Framework for School Education 2023 is the spine, and AI from Class 3 is the most visible add-on. Schools that have spent the last year preparing administratively will find this manageable. Schools that left it to the last term will spend 2026-27 catching up.

The next milestones to watch are NCERT's release of the final curriculum framework and the first NISHTHA module roster, expected in batches through the summer.