CBSE CT & AI Class 7 Syllabus 2026–27: AI Domains, ML Concepts and Bias

Class 7 is where CBSE’s AI content stops being introductory. Classification, regression and clustering arrive alongside Computer Vision and Natural Language Processing — and bias awareness arrives with them rather than after. The allocation stays at 100 hours: 40 advanced computational thinking, 20 introductory AI, 40 interdisciplinary projects.

Teaching fairness at the same moment as prediction is a deliberate sequencing choice, and a good one. Students meet the idea that a model can be confidently, systematically wrong before they have had a year to assume otherwise.

Part 1 — Computational Thinking chapters

1. Large Numbers Around Us — multi-digit numbers and place value
2. Arithmetic Expressions — operator precedence and calculation methods
3. A Peek Beyond the Point — decimals and binary representation
4. Expressions using Letter-Numbers — algebraic thinking with variables
5. Parallel and Intersecting Lines — geometric relationships and spatial reasoning
6. Number Play — logic puzzles and systematic problem-solving
7. A Tale of Three Intersecting Lines — triangle properties and geometric analysis
8. Working with Fractions — fractional and proportional reasoning

A Peek Beyond the Point is the sleeper chapter. Binary representation is the first time students see that everything a machine handles is ultimately numbers — which makes “how does a computer see a photograph?” answerable in the Computer Vision chapter rather than magical.

Part 2 — the four AI chapters

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

Teaching classification, regression and clustering to twelve-year-olds

These three words intimidate teachers more than students. The distinction is simpler than the vocabulary suggests, and it can be taught entirely without software.

Concept	The question it answers	An unplugged activity
Classification	Which category does this belong to?	Sort 40 paper slips into two labelled bins using a rule the class writes. Then find a slip the rule gets wrong.
Regression	What number should I predict?	Plot classmates’ height against arm span, draw a line by eye, predict a missing value, then measure it.
Clustering	What natural groups exist that nobody told me about?	Give groups unlabelled data about lunchboxes or travel times and ask them to find groupings. Compare across groups — different answers are the lesson.

The critical difference to draw out: classification and regression are told the right answer during learning; clustering is not. Students who grasp that distinction have understood supervised versus unsupervised learning without needing the terms.

Computer Vision and NLP without a lab

Both are teachable on paper. For Computer Vision, give students a grid of squares with numbers for darkness and ask them to write a rule that finds an edge — that is genuinely how early vision works, and it connects directly to binary representation from the CT half. For NLP, have them write rules to decide whether a sentence is positive or negative, then find sentences that break every rule they wrote. Sarcasm defeats them quickly, which is exactly the point.

The bias chapter is the most important one

Ethics and AI Bias Awareness is where the year’s content becomes durable. The mistake is teaching it as a warning. Teach it as a prediction: once students have built classification rules themselves, ask who does your rule treat unfairly, and why? Every rule a class invents will disadvantage somebody, and discovering that in their own work lands far harder than hearing it about someone else’s algorithm.

Link it back to the data chapter: bias is usually the training data’s history showing through, not anyone’s intent. That framing is what the Class 8 chapter Data and Fairness in AI builds on next year.

What teachers get wrong at Class 7

• Defining the three ML terms and moving on. Definitions without the sorting activity do not survive to Class 8.
• Skipping binary. A Peek Beyond the Point makes Computer Vision explicable. Teach it first.
• Treating AI in Industries as a list. Four sectors read out is forgettable; one sector examined properly — what data, what decision, what could go wrong — is not.
• Leaving bias to the last week. It works best interleaved with the ML activities, not appended to them.

Frequently asked questions

What AI topics are in CBSE Class 7?

Four chapters: AI Domains and Applications (Data Science, Computer Vision, NLP, plus classification, regression and clustering); AI in Industries; Data Visualisation and Analysis; and Ethics and AI Bias Awareness.

Does Class 7 require students to build machine learning models?

No. Classification, regression and clustering are taught as concepts and applications. The 20-hour allocation makes technical model-building neither expected nor feasible.

How many hours is Class 7 CT & AI?

One hundred hours a year: 40 advanced computational thinking, 20 introductory AI, and 40 interdisciplinary projects delivered as two 20-hour projects.

Can Class 7 AI be taught without computers?

Largely yes. Classification, clustering, edge detection and rule-based sentiment analysis all work unplugged. Devices help most with Data Visualisation and Analysis, where a spreadsheet saves considerable time.

What is the difference between classification and clustering?

Classification sorts items into categories you already defined; clustering finds groupings nobody specified in advance. Classification is told the right answer while learning; clustering is not.

How should Class 7 AI be assessed?

Continuously and competency-based — projects, reflective journals, group discussion and teacher observation alongside written and practical work. A student who can explain who their own classification rule treats unfairly has demonstrated more than any multiple-choice test would show.

---

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

Related: CBSE AI curriculum overview · Class 6 · Class 8 · AI glossary

Source: CBSE Computational Thinking & AI Class 7 handbooks (2026–27), cbseacademic.nic.in.