CBSE AI Curriculum 2026–27: What Every School Must Do (Classes 3–8)

From the 2026–27 session, CBSE’s Computational Thinking and Artificial Intelligence curriculum applies to Classes 3 to 8 — 50 hours a year at the preparatory stage (Classes 3–5) and 100 hours across the middle stage (Classes 6–8), split as 40 hours advanced computational thinking, 20 hours introductory AI and 40 hours of interdisciplinary projects. CBSE has published student and teacher handbooks for every class.

This guide sets out the actual chapter list for each class, how the hours break down, what the assessment model expects, and the decisions a school has to make before the term starts. It is written from the published CBSE handbooks rather than from press coverage.

The structure at a glance

Stage	Classes	Hours	How it is delivered
Preparatory	3, 4, 5	50 hours per year	Integrated into existing Mathematics and The World Around Us periods — not a separate subject
Middle	6, 7, 8	100 hours per year	40 hours advanced CT + 20 hours introductory AI + 40 hours interdisciplinary projects (two projects of 20 hours each)

The first thing to notice is that at the preparatory stage this is not a new period on the timetable. It is embedded in Maths and TWAU, which means the class teacher usually delivers it rather than the computer teacher — a point many schools get wrong when assigning staff.

The second is that AI proper is only 20 hours a year, and only from Class 6. Everything before that is computational thinking. Schools that rush to buy AI tools before building CT foundations are solving the wrong problem.

Classes 3–5: computational thinking through Maths

The preparatory stage targets four competencies — abstract thinking, pattern recognition, decomposition and algorithmic thinking — plus analytical and verbal reasoning, taught entirely through activity-based chapters with familiar contexts. There is no AI content at this stage and no requirement for devices.

The chapter titles give the flavour better than any description. Class 3 includes What’s in a Name?, Toy Joy, Vacation with My Nani Maa, Fun with Shapes, Raksha Bandhan, Fair Share, Filling and Lifting, Time Goes On and The Surajkund Fair. Class 4 moves to Shapes Around Us, Hide and Seek, Patterns Around Us, Thousands Around Us, Equal Groups, Fun with Symmetry, The Transport Museum and Data Handling. Class 5 covers We the Travellers, Angles as Turns, The Dairy Farm, Shapes and Patterns, Coconut Farm, Grandmother’s Quilt, Racing Seconds, Animal Jumps, Maps and Locations and Data Through Pictures.

These read like mathematics chapters because they largely are. The computational thinking is in how the problems are posed — pathfinding in Hide and Seek is algorithmic thinking, Grandmother’s Quilt is pattern recognition, Fair Share is decomposition. A teacher who does not see that will teach it as ordinary maths and miss the point entirely, which is precisely why teacher preparation matters more than resources at this stage.

Class 6: introduction to AI

Class 6 is where AI enters. The handbook is in two parts.

Part 1 — Computational Thinking runs through the mathematics chapters: Patterns in Mathematics, Lines and Angles, Number Play, Data Handling and Presentation, Prime Time, Perimeter and Area, Fractions, Playing with Constructions, Symmetry and The Other Side of Zero.

Part 2 — Artificial Intelligence has four chapters:

Chapter	Topics
Introduction to AI and Everyday Examples	Meaning of AI, AI in daily life, AI and automation, human vs machine intelligence, types of learning in AI
Basic Data Concepts	Understanding data, types of data, collecting data, organising data, representing data
Simple Pattern Recognition and Decision Making	Understanding patterns, identifying patterns, observations and conclusions, decision making
Ethics and Digital Responsibility	Responsible use of technology, online safety, privacy, password safety, digital footprints

Note what is absent: no coding, no model building, no tools. Class 6 AI is conceptual and ethical. A school that responds to this by buying a coding platform has misread the curriculum.

Class 7: AI domains and bias

Part 1 — Computational Thinking: Large Numbers Around Us, Arithmetic Expressions, A Peek Beyond the Point (which introduces binary representation), Expressions using Letter-Numbers, Parallel and Intersecting Lines, Number Play, A Tale of Three Intersecting Lines and Working with Fractions.

Part 2 — Artificial Intelligence:

Chapter	Topics
AI Domains and Applications	Data Science, Computer Vision and Natural Language Processing; classification, regression and clustering
AI in Industries	Applications across healthcare, education, transport and communication
Data Visualisation and Analysis	Collecting, organising and interpreting data through visual representation
Ethics and AI Bias Awareness	Responsible AI use, fairness, digital citizenship

Class 7 is the step up. Classification, regression and clustering are real machine-learning concepts, and bias awareness arrives alongside them rather than after — a sound sequencing decision, since students meet fairness at the same moment they meet prediction.

Class 8: the AI project lifecycle

Part 1 — Computational Thinking: A Square and a Cube, Power Play, A Story of Numbers, Quadrilaterals, Number Play, We Distribute Yet Things Multiply and Proportional Reasoning.

Part 2 — Artificial Intelligence:

Chapter	Ethical focus
AI Project Lifecycle	Responsible problem solving
Artificial Intelligence and Its Applications	Understanding the impact of AI systems
Data and Fairness in AI	Promoting fairness and inclusivity
Ethics and Responsible AI	Ethical decision-making in technology

Every Class 8 AI chapter carries an explicit ethical dimension. The AI Project Lifecycle — define the problem, collect data, test AI tools, reflect and improve — is the spine of the year, and the two 20-hour interdisciplinary projects are where students actually run it.

Class-by-class syllabus guides

Each class has its own chapter list, hours and teaching pitfalls. Full breakdowns:

• Class 3 syllabus — 14 CT chapters, 50 hours, no AI, no devices
• Class 4 syllabus — 14 CT chapters including Hide and Seek and Data Handling
• Class 5 syllabus — 15 CT chapters; the bridge year into Class 6 AI
• Class 6 syllabus — AI enters: 4 AI chapters, 20 hours
• Class 7 syllabus — classification, regression, clustering, CV, NLP and bias
• Class 8 syllabus — the AI Project Lifecycle and responsible AI

How assessment works

CBSE’s Class 8 handbook is explicit that assessment should be continuous, formative and competency-based, focusing on the ability to apply knowledge rather than on memorisation. In practice the modes span written tests, practical exams, thematic projects, reflective journals, group discussions and teacher observation journals.

Two practical consequences most schools discover too late:

• Reflective journals cannot be backfilled. If they start in January they are worthless. They have to run from week one.
• Projects need checkpoints designed into the calendar. Two 20-hour projects at the middle stage is a substantial commitment, and retrofitting them into the final month is the commonest failure mode.

What a school must actually decide

1. Who teaches it. At Classes 3–5 the content sits inside Maths and TWAU, so the class teacher is usually the right choice — not automatically the ICT teacher.
2. Where the hours live. Fifty and 100 hours have to appear in a written timetable, not a plan. Hours that exist only in a plan become a fraction of themselves by February.
3. What resources you use. CBSE publishes student and teacher handbooks per class. Anything you buy on top should map chapter-for-chapter onto them, or your teachers end up doing the mapping.
4. How you assess. Rubrics and project checkpoints before term, not during it.
5. What your AI policy says. Five lines will do to start: what is allowed, what is banned, what must be disclosed, what data never goes into an AI tool, and who decides exceptions.

You can score your school against all five areas with our free school AI readiness self-check. The full implementation sequence is in AI literacy for schools.

The teacher problem

The curriculum is not the hard part — it is published, structured and reasonable. The hard part is that several lakh teachers now have to teach content most were never trained in, in a subject where being one week ahead of the class does not work because students ask questions the handbook does not answer.

Nationally this is being addressed at scale: the Ministry of Education’s AI Literacy for Teachers programme, launched in May 2026 through Bodhan AI at IIT Madras, targets over a million teachers by 2027. At school level the evidence favours starting with the adults — Microsoft’s 2026 research found organisational factors accounted for 67% of AI impact against 32% for individual effort, and that value rose 17 points where managers modelled use rather than mandating it.

Related: AI literacy for teachers · CT & AI teacher training for CBSE · CT & AI curriculum resources

Frequently asked questions

Which classes does the CBSE AI curriculum apply to in 2026–27?

Classes 3 to 8. Classes 3–5 get 50 hours a year of computational thinking embedded in Mathematics and The World Around Us. Classes 6–8 get 100 hours covering advanced CT, introductory AI and interdisciplinary projects.

Is CT & AI a separate subject with its own exam?

At the preparatory stage it is integrated into existing Maths and TWAU periods rather than taught separately. Assessment is described as continuous, formative and competency-based rather than a single terminal examination. Check current CBSE Academic circulars for the reporting requirements that apply to your school.

Where can I download the CBSE CT & AI handbooks?

CBSE publishes student and teacher handbooks per class on the CBSE Academic website under the 2026–27 curriculum materials. There is a separate student handbook and teacher handbook for each class from 3 to 8.

How much AI is actually in the CBSE AI curriculum?

Twenty hours a year, from Class 6 onward. Classes 3–5 have no AI content at all — they build computational thinking. The name causes confusion; the weighting is heavily toward CT.

Do students need laptops or tablets for this?

Not for Classes 3–5, which is activity-based and unplugged. From Class 6 devices help for data handling and visualisation, but a large share of the AI content — patterns, decision making, ethics, bias — works perfectly well on paper.

Does the CT & AI curriculum include coding?

Not as a stated requirement in the Classes 3–8 AI chapters. Algorithmic thinking is taught throughout, and Class 7 touches binary representation, but the AI content is conceptual, applied and ethical rather than programming-based.

When does this take effect?

The 2026–27 academic session. Handbooks for all six classes are already published, so preparation can begin immediately.

---

PiyushAI provides chapter-exact CT & AI curriculum resources for Classes 3–8, teacher training aligned to the CBSE handbooks, and the PiyushAI School Lab. Founded by Piyush Wairale (IIT Madras).

Related: complete AI literacy guide · AI literacy for schools · school readiness self-check · AI glossary

Sources: CBSE Computational Thinking & AI student and teacher handbooks for Classes 3, 4, 5, 6, 7 and 8 (2026–27), published on cbseacademic.nic.in; Microsoft 2026 Work Trend Index; AI Literacy for Teachers programme.