Artificial intelligence in CBSE schools, 2026-27

What CBSE launched in April 2026

In circular Acad-15/2026, dated 1 April 2026, CBSE launched its Computational Thinking and Artificial Intelligence curriculum, called CT and AI here, for Classes 3 to 8 from the 2026-27 session. The circular says the aim is to develop "AI-ready learners" through logical reasoning, problem-solving, pattern recognition and algorithmic thinking, alongside digital literacy and the ethical use of technology. It tells schools to take the steps needed to put the curriculum in place.

A second circular, Acad-18/2026 of 9 April, adds two things for schools. The target is artificial intelligence literate learners by 2030. And from 2026-27, the other artificial intelligence courses CBSE schools ran until 2025-26 are discontinued for Class IX. Class X stays on the 2025-26 scheme this year, and NCERT is to supply CT and AI modules for Classes 9 to 12, taught as modules for internal assessment.

Computational thinking comes first

CBSE's framework treats computational thinking, the habit of breaking a problem into smaller steps and spotting patterns, as the foundation for learning artificial intelligence. Artificial intelligence is introduced later, once that groundwork is in place.

The suggested time is 50 hours a year in Classes 3 to 5 and 100 hours a year in Classes 6 to 8. The framework calls this flexible, and schools may go beyond it. In Classes 3 to 5 there is no new subject: computational thinking is built into Mathematics and The World Around Us through extra questions and activities in resource books.

In Classes 6 to 8 the hours split three ways, and only one share teaches artificial intelligence directly.

That is one hour in five. The rest builds the reasoning skills the framework says artificial intelligence depends on, and applies them in cross-subject projects.

Who teaches it

In Classes 3 to 5, subject teachers deliver the curriculum inside their ordinary lessons, using resource books that follow the same chapters as the textbook. The handbooks circular says they are not textbooks but companions to the NCERT Mathematics books, so a teacher can use the matching questions as each concept is taught.

In Classes 6 to 8, subject teachers cover computational thinking and computer teachers cover artificial intelligence literacy. Projects that combine the two are assessed by the computer teacher.

CBSE asks teachers to read each chapter's questions first and then to guide rather than give answers. Students get time to attempt a problem alone, and then the class compares approaches.

How it will be assessed

The framework moves away from traditional summative assessment towards continuous, formative and competency-based assessment. It says the focus is the ability to apply knowledge, creativity and ethical reasoning rather than rote memorisation.

In Classes 3 to 5 the suggested methods are written tests with computational thinking questions and puzzles, group activities and a Teacher Observation Journal. Classes 6 to 8 add practical examinations, thematic projects and reflections. Teachers are expected to write clear, consistent rubrics.

This changes what a marking day looks like. The handbooks circular says the process of thinking matters more than reaching the correct answer, so a teacher has to look at how a student got there.

Ethics is part of the syllabus

The framework includes ethics from Class 6. Class 6 covers online safety, privacy and digital footprints. Class 7 introduces bias in intelligent systems, and Class 8 covers data and fairness, privacy and misinformation. Class 8 students also try simple no-code tools such as image classifiers and chatbots.

The framework says regular ethics teaching helps students verify information, detect bias and notice when reliance on these tools becomes what it calls "cognitive over-reliance". For what the research says about chatbots as tutors, see do tutoring chatbots help students learn?

What came before: artificial intelligence across subjects

The 2026 curriculum follows an earlier CBSE approach. CBSE's Artificial Intelligence Integration manual, written after NEP 2020, said CBSE schools should begin integrating artificial intelligence with other disciplines from Classes 1 to 12. It proposed a three-step "AI+X" method: choose a topic from your own subject (the X), find an artificial intelligence concept that shares ground with it, and attach the two in the lesson plan, with colleagues or the computer teacher.

Its example pairs data collection in Mathematics with data acquisition in artificial intelligence. The manual was built with CBSE teachers in webinars, and subject manuals followed for Hindi, English, Mathematics, Social Science and Science.

The 2026 framework keeps the cross-subject idea, saying computational thinking should be built into subjects beyond Mathematics. What is new is that it comes with hours, resource books and assessment methods.

Where the gap between policy and practice shows

The framework says what to teach, who teaches it and how long to spend. It says less about how a school gets ready. Circular Acad-15/2026 leaves that to schools. The framework document does not mention teacher training at all, and the nearest guidance is the handbooks circular's advice on using the resource books.

Equipment is the other question. The framework says the curriculum does not depend on any platform and recommends free and open-source tools, so Classes 3 to 5 need worksheets rather than software. Class 6 to 8 projects and no-code tools do need devices, and the national figures show how much of that is in place. The Ministry of Education's UDISE+ 2025-26 report counts 14.67 lakh schools. 69.9% had computers and 67.4% had internet, up from 64.7% and 63.5% in 2024-25. That leaves roughly one school in three with no computer, and about the same share with no internet connection. CBSE schools are one part of that total and the report summary does not split them out, so check your own school's position rather than reading it off a national average.

What a school can take from this

• Name an owner for each part. In Classes 3 to 5, that is the Mathematics and World Around Us teachers. In Classes 6 to 8, it is the subject teachers and the computer teacher, who also assesses the joint projects.
• Put the resource books beside the schemes of work. Follow the handbooks circular: teachers read the questions for a chapter before teaching it.
• Write rubrics before the first project. The framework asks for clear, consistent rubrics in Classes 6 to 8, and it is easier to agree them before work is marked.
• Keep a record of hours. The 50 and 100 hour figures are a guide, but a log shows where the time is going.
• Check Class IX. If your school ran a separate artificial intelligence course, it is discontinued for Class IX from 2026-27.
• List what the Class 6 to 8 projects need. Count working computers and the connection, then plan sessions around them. Free and open-source tools are encouraged.
• Plan staff time on the ethics topics. Bias, privacy and checking what a tool produces are part of the syllabus, and teachers will be asked about them.