ISC Class 12 Artificial Intelligence Syllabus 2026-27

The ISC (Indian School Certificate) Class 12 Artificial Intelligence syllabus for the academic session 2026-27 is prescribed by the Council for the Indian School Certificate Examinations (CISCE) under subject code 883. Artificial Intelligence is one of the newest subjects on the ISC curriculum: CISCE introduced it for Classes 11 and 12 from the 2025-26 session, which makes the February-March 2027 examination the first full ISC board examination in the subject.

The Class 12 course assumes the Class 11 groundwork in Python, matrices, statistics and data visualisation, and moves on to applied work: named entity recognition and text summarisation, hypothesis testing, artificial neural networks, machine learning in real industry settings, and computer vision with OpenCV. It is assessed across two components — a three-hour theory paper worth 70 marks and a 30-mark practical component made up of a practical examination and a practical file.

This page sets out the complete unit-wise syllabus with sub-topics, the official marks weightage for each unit, the exam pattern, the practical assessment breakdown, and preparation guidance for ISC 2027 candidates.

Note: Artificial Intelligence may be taken with Computer Science, but not with Robotics. The two are separate ISC electives with separate syllabuses and separate papers.

Quick Facts: ISC Class 12 Artificial Intelligence 2026-27

ISC Class 12 Artificial Intelligence Exam Pattern 2026-27

The Artificial Intelligence examination follows a two-paper format. Paper I is the written theory examination set and marked by CISCE. Paper II is the practical component, evaluated by a Visiting Examiner appointed locally and approved by CISCE.

The practical examination itself is split into two sittings within the same three hours. Candidates get a maximum of 90 minutes for a Planning/Writing session, in which they prepare an algorithm and a handwritten program, and then 90 minutes for the Examination session, in which they code, execute and demonstrate the program on a computer. Candidates may only move to the Examination session once the full 90 minutes of planning are over, and a large gap between the planned program and the final printout costs marks.

ISC Class 12 Artificial Intelligence Syllabus 2026-27: Detailed Unit-Wise Breakdown

The Class 12 theory syllabus is organised into five units. Key concepts from the Class 11 course are treated as a prerequisite and need to be revised, even though the Class 12 paper is set only on the Class 12 portion.

Unit 1: Applications of AI (10 Marks)

This unit applies AI techniques to language, data and inference. It is the bridge between the Class 11 foundations and the modelling units that follow.

1.1 Natural Language Processing (NLP)

• Named Entity Recognition (NER): identifying and classifying names of people, organisations and locations in a text such as a news article

• Text summarisation using extraction methods: condensing a long passage by pulling out its most important sentences or phrases

• Spam email detection using machine learning: building a classifier that separates spam from non-spam on the basis of content and features

1.2 Statistics and Probability

• Reading data trends using mean, median, mode, range, variance and standard deviation

• Probability as a measure of uncertainty

• Common distributions: uniform, normal and binomial

1.3 Hypothesis Testing

• Null and alternative hypotheses and their relevance to AI experiments

• Type I and Type II errors

• Regression analysis: simple linear regression for prediction

• Correlation coefficient as a measure of the relationship between two variables

Unit 2: Different Paradigms of AI — Neural Networks, Machine Learning, Deep Learning (15 Marks)

This unit introduces the three major modelling paradigms and the vocabulary used throughout the rest of the course.

2.1 Artificial Neural Networks (ANN)

• What an artificial neural network is; nodes (neurons)

• Input layer, hidden layer and output layer

• Forward propagation and backward propagation

• The ANN model and the multi-layer perceptron (MLP)

• Worked example: recognising handwritten digits using an MLP

2.2 Machine Learning (ML)

• Machine learning as the AI paradigm that lets computers learn from data and make predictions or decisions without being explicitly programmed

• Worked example: predicting house prices from features such as size, location and number of bedrooms using a decision tree

2.3 Deep Learning (DL)

• What deep learning is, and how neural networks are used for image recognition, pattern recognition and computer vision

• Worked example: classifying images into categories (cat, dog, bird) using a Convolutional Neural Network (CNN)

Unit 3: Practical Implications of ANN (15 Marks)

The most hands-on unit of the theory paper. It follows the full modelling workflow from raw data to a tuned network, and questions here often involve reading Python code or interpreting evaluation metrics.

3.1 Data Preparation

• NumPy for numerical computation and Pandas for data manipulation

• Scikit-learn for preprocessing: scaling, encoding categorical variables, and splitting data into training and testing sets

3.2 Building the Neural Network

• Choosing between Keras and TensorFlow, and why Keras (which runs on top of TensorFlow) is the beginner-friendly option

3.3 Training the Neural Network

• Fitting data to the model using the fit() function

• Specifying the number of epochs and the batch size

3.4 Model Evaluation

• Evaluating performance on test data using accuracy, precision and recall

• Visualising performance metrics with libraries such as Matplotlib

3.5 Tuning

• Grid search and random search for hyperparameter tuning

• Tuning the learning rate, batch size, and the number of hidden layers and neurons

Unit 4: Practical Implications of Machine Learning (20 Marks)

The highest-weighted unit in the paper. It is case-study driven: each application should be learned with its setting, its data, and the reason a particular model suits it.

4.1 Predictive Maintenance

• Using historical machine data to predict equipment failure or the need for maintenance

• Minimising downtime and optimising maintenance schedules

4.2 Fraud Detection

• Detecting credit card fraud and identity theft

• How financial institutions use these models to act before losses occur

4.3 Medical Diagnosis

• Analysing medical imaging, genetic information and patient records to support diagnosis

Unit 5: Introduction to Computer Vision (10 Marks)

The closing unit covers OpenCV and the standard computer-vision tasks, each with a real-world illustration.

5.1 Introduction to CV and OpenCV in Real-World Scenarios

• Image classification, object detection, facial recognition, colour detection and object tracking in brief, with examples

5.2 Image Classification

• Implementing a simple classifier that distinguishes between categories of objects, such as cats against dogs or fruits against vegetables

5.3 Object Detection

• Identifying and localising specific objects within an image, such as cars in a street scene or faces in a photograph

5.4 Facial Recognition

• Building an application that recognises and identifies faces in images or video, and optionally associates them with known individuals

5.5 Colour Detection

• Identifying and classifying objects by their colour properties, such as sorting coloured objects on a conveyor belt or spotting ripe fruit

5.6 Object Tracking

• Tracking the movement of objects through a video sequence, such as following a ball in sports footage or monitoring vehicles in traffic surveillance

Unit-Wise Overview Table

The table below summarises all five units of the ISC Class 12 Artificial Intelligence 2026-27 syllabus with their official weightage in Paper I.

ISC Class 12 Artificial Intelligence Marking Scheme 2026-27

The 70 marks of Paper I are supplemented by a 30-mark practical component. Candidates are expected to complete a minimum of 25 assignments across the year — 15 practical programming assignments and 10 hands-on case studies.

Within the 15-mark practical examination, marks are credited for the algorithm (3 marks), the Python program with internal documentation (7 marks), the hard-copy printout (2 marks) and the output (3 marks). Across the whole subject, candidates must secure a minimum of 35% to pass, in line with the ISC passing criterion for each individual subject.

Laboratory Requirements for ISC Class 12 Artificial Intelligence

Schools offering the subject are expected to provide a desktop and webcam for every pair of students, running Windows or Linux with Python installed. The prescribed software stack is worth knowing by name, because the theory paper draws its code questions from it:

• IDEs: Jupyter Notebook or Google Colab for cloud-based work; PyCharm or VS Code for offline programming

• Core libraries: NumPy, Pandas and Matplotlib for data manipulation and visualisation

• Machine learning: Scikit-learn

• Deep learning: TensorFlow or PyTorch

• Computer vision: OpenCV

• NLP: NLTK or spaCy

• Databases: MySQL or SQLite

• Version control and cloud: GitHub, Google Colab or AWS

Preparation Tips for ISC Class 12 Artificial Intelligence 2026-27

Artificial Intelligence rewards students who can actually run the code they write about. Because 2027 is the first board examination in the subject, there is no bank of past papers to fall back on, which makes syllabus discipline and practice papers more important than usual.

Dry-Run Every Python Snippet

A significant share of theory questions ask for the output of a short program. Work line by line through NumPy, Pandas and scikit-learn snippets and write the output exactly as it would print, including data types and array shapes. Guessing the shape of a result is the single most common way students lose output marks.

Learn the Evaluation Formulas Cold

Accuracy, precision and recall come up repeatedly, and so do parameter counts for Dense layers and Gini impurity for decision trees. Practise computing them by hand from a confusion matrix until the arithmetic is automatic, and remember to include the bias term when counting parameters.

Treat Unit 4 as Case Studies, Not Definitions

Predictive maintenance, fraud detection and medical diagnosis carry 20 marks between them — the largest block in the paper. Each answer should be tied to its setting: the factory floor, the bank, the hospital. A generic description of machine learning will not earn the application marks. It is also worth remembering why accuracy alone is a poor metric for rare events such as fraud or machine failure.

Practise Forward Passes by Hand

For small networks, work through a forward pass and a single backpropagation weight update on paper. This is the fastest way to make layers, weights and activation functions concrete, and it makes diagram-based questions on ANN structure straightforward.

Keep the Practical File Current

The practical file carries 15 marks between internal evaluation and the Visiting Examiner — the same as the practical examination itself. Twenty-five assignments is a lot to reconstruct in February, so complete and document them as you go, with clear algorithms and printed outputs.

Watch the OpenCV Details

Computer vision questions often turn on specifics: OpenCV stores colour images in BGR order rather than RGB, and hue values run from 0 to 179 rather than 0 to 359. Similarly, a scaler must be fitted after the train-test split, not before. These are small points that examiners use to separate confident candidates from the rest.

Solve Full-Length Specimen Papers Under Timed Conditions

Once the syllabus is complete, sit a full paper in one go with 15 minutes of reading time, then mark it against the answer key. You can download both ISC Class 12 Artificial Intelligence specimen papers with full answer keys to check your timing and your coverage.

Frequently Asked Questions (FAQs)

What is the subject code for ISC Class 12 Artificial Intelligence?

The subject code for ISC Class 12 Artificial Intelligence is 883, as assigned by the Council for the Indian School Certificate Examinations (CISCE).

How many marks is the ISC Class 12 Artificial Intelligence exam?

The subject carries 100 marks in total: 70 marks for Paper I (Theory) and 30 marks for Paper II (Practical). The theory paper is three hours long, with an additional 15 minutes of reading time.

How many units are there in the ISC Class 12 Artificial Intelligence syllabus?

There are five units: Applications of AI (10 marks), Different Paradigms of AI (15 marks), Practical Implications of ANN (15 marks), Practical Implications of Machine Learning (20 marks) and Introduction to Computer Vision (10 marks).

Which unit carries the highest weightage?

Unit 4, Practical Implications of Machine Learning, carries the highest weightage at 20 marks. It covers predictive maintenance, fraud detection and medical diagnosis.

Which programming language is used in ISC Class 12 Artificial Intelligence?

Python. The prescribed libraries include NumPy, Pandas, Matplotlib, scikit-learn, Keras or TensorFlow, OpenCV, and NLTK or spaCy. The theory paper can ask for the output of a program or for short pieces of code.

Can I take Artificial Intelligence and Robotics together at ISC?

No. CISCE offers Artificial Intelligence (883) and Robotics (884) as separate electives, and the syllabus states that Artificial Intelligence may be taken with Computer Science but not with Robotics.

Is Class 11 Artificial Intelligence needed for the Class 12 paper?

The ISC Class 12 examination paper is set only on the Class 12 portion of the syllabus. However, the syllabus explicitly notes that key Class 11 concepts — Python programming, matrices, descriptive statistics and data visualisation — need to be revised as a prerequisite, because the Class 12 units build directly on them.

How many practical assignments are required?

Candidates are expected to complete a minimum of 25 assignments during the year: 15 practical programming assignments and 10 hands-on case studies, all recorded in the practical file.

What is the passing mark for ISC Class 12 Artificial Intelligence?

Candidates must secure a minimum of 35% in the subject, which is the ISC passing criterion applied to each individual subject.

Where can I download the official ISC Class 12 Artificial Intelligence syllabus?

The official syllabus PDF for Artificial Intelligence (883) can be downloaded from the CISCE website at www.cisce.org, under the Publications or Syllabus section for the relevant ISC examination year.