Explore the CBSE AI syllabus Class 10 for 2026-27: all 7 units, 100-mark split, practical file rules and smart prep tips from SSD Prayas.
Ask a Class 10 student what AI is and you'll probably hear "ChatGPT." Ask what the CBSE AI syllabus for Class 10 actually expects from them, and the answer gets vague fast. That gap is worth closing early. Code 417 isn't a subject you can cram the night before, because half its marks come from things you build, run and explain out loud. So, here's the 2026-27 syllabus, unit by unit, with the marks split and a realistic way to prepare.
What Is CBSE Class 10 AI (Code 417)?
Artificial Intelligence, subject code 417, is a skill subject CBSE offers in Classes 9 and 10. It's optional, so each school decides whether to run it. The course carries 100 marks: 50 for a two-hour written paper and 50 for practical and project work.
The goals are down to earth. Students meet the three domains of AI, learn the AI project cycle, pick up basic Python, and try building small solutions for real social problems. Nobody expects a Class 10 student to train a giant model. They're expected to think clearly about data, which is harder than it sounds.
Why It's Worth Taking Seriously
Board marks aside, this subject builds habits that travel well. Asking what the data really says. Checking whether a result holds up. Wondering who a system might treat unfairly. Those habits help whether a student heads towards science, commerce or something nobody has named yet.
How the 100 Marks Are Divided
| Component | Marks |
|---|---|
| Written paper (2 hours) | 50 |
| Practical file (minimum 15 programs) | 15 |
| Practical exam | 15 |
| Viva voce | 5 |
| Project, field visit or portfolio | 10 |
| Viva on the project | 5 |
| Total | 100 |
The written paper itself splits into 10 marks of Employability Skills (Part A) and 40 marks from subject-specific AI skills (Part B). Notice how much the practicals weigh. A student who ignores the practical file and hopes to recover in theory is giving away half the subject.
The 2026-27 Class 10 AI Syllabus, Unit by Unit
Part B has seven units. Here's what each one really involves.
1. Revisiting the AI Project Cycle and Ethical Frameworks. Problem scoping, data acquisition, exploration, modelling, evaluation: the same loop working data teams follow. This unit also brings in ethics, through ethical frameworks and a bioethics case study. Don't skim it. Case-based questions love this material, and it's where a student can show real thinking instead of memorised lines.
2. Advanced Concepts of Modelling in AI. The engine room of the syllabus. It covers how machines learn patterns from data, and it's what makes the later units click. Spend time on the intuition before the vocabulary.
3. Evaluating Models. Building a model is half the job. Knowing whether it's any good is the other half. Expect the confusion matrix, accuracy, precision, recall and F1 score. Working these out by hand a few times beats reading about them ten times.
4. Statistical Data. Practical-heavy and tested in the practical exam. Students work with tools like Orange Data Mining and MS Excel to analyse data and read what it's telling them.
5. Computer Vision. How machines make sense of images. It's the most visual unit, so results show up on screen, which makes practice more satisfying than most.
6. Natural Language Processing. Chatbots, the difference between human and computer language, and text processing with Bag of Words, TF-IDF and NLTK. This is where many students realise a chatbot isn't "understanding" anything the way we do.
7. Advance Python. The smallest unit by hours, roughly 10, but it feeds straight into the practical exam. Shaky Python shows up everywhere else.
Alongside these sits Part A: five Employability Skills units covering communication, self-management, ICT, entrepreneurial and green skills, at 2 marks each. Easy marks if you actually read them.
The Practical Side: Where 50 Marks Live
The practical file needs at least 15 programs. The practical exam draws on Statistical Data, Computer Vision, NLP and Advance Python, and comes with a short viva. Then there's the project, field visit or portfolio, followed by its own viva.
The project is the part students underestimate. Pick a problem you can see from your own street or school, like waste segregation, bus timings, or a library that never has the right books, and scope it properly. A viva makes it obvious whether you built the thing or borrowed it.
How to Prepare Without Burning Out
- Start with the official CBSE curriculum PDF for the session. Notes and sample papers vary in quality; the PDF is your checklist.
- Write programs weekly. Fifteen programs squeezed into the last fortnight look exactly like that.
- Practise evaluation numbers on paper. Matrix, precision, recall, F1. Repeat until it's boring.
- Explain each unit out loud to someone who doesn't know it. Where you stall is where the gap is.
- Read the ethics case studies twice, then argue both sides.
Where SSD Prayas Fits In
Reading a syllabus is one thing. Sitting with a mentor while your first model misbehaves is another. SSD Prayas runs AI skilling programmes for school students from Class 3 to 12, along with training for educators and professionals, and works with schools and partners so students learn by doing. For a Class 10 student, that kind of practice makes the practical file, the project and the viva feel familiar instead of frantic. Schools and parents can explore current programmes at ssdprayas.com.
FAQs
Is AI compulsory in Class 10 CBSE?
No. Code 417 is an optional skill subject, so it depends on whether your school offers it. CBSE has separately announced a Computational Thinking and AI subject for younger classes from 2026-27, so keep an eye on official circulars.
Do I need to know coding?
Some, yes. Basic Python is part of the course, and Advance Python is a full unit tested in practicals. Starting from zero is fine.
How many marks is the written paper?
Fifty, in two hours: 10 from Employability Skills and 40 from the subject-specific units.
Final Thoughts
The CBSE AI syllabus for Class 10 rewards students who do things: run the model, break it, fix it, explain it. Treat the practical file and project as the main event, not the paperwork, and the theory paper gets much easier. Start now, work in small weekly chunks, and use the official curriculum as your checklist.
