From 2027, MOE will make the refreshed Code for Fun programme available to all schools as part of its push to strengthen students’ AI literacy. In secondary schools, Code for Fun and AI for Fun help pupils understand computational thinking, digital making and responsible use of emerging technology. Under Full Subject-Based Banding (Full SBB), academic G1, G2 and G3 levels are not permanent labels of who can be creative with technology. What a student actually attends depends on the receiving school’s plan, project level and approved participation arrangements.

For Punggol families, the practical question is what a teenager will build, test and explain rather than which software has the most impressive name. IMDA’s earlier secondary Code for Fun framework includes microcontrollers, design prototyping, block-based or text-based programming, and hands-on problem solving. MOE’s March 2026 announcements confirm an AI-focused update for 2027. Parents should verify the actual school’s implementation rather than assume every pupil will study Python, receive a robot or sit an AI examination on the same date.

• MOE announced in March 2026 that refreshed Code for Fun would be made available to all schools in 2027; actual school implementation must be confirmed.
• The existing secondary Code for Fun framework developed by MOE and IMDA includes design thinking, digital making, computational thinking and microcontroller projects.
• IMDA’s published secondary framework described a 10-hour programme with block-based and text-based tracks; 2027 updates may alter activities and duration.
• AI for Fun modules have been available to government and government-aided primary and secondary schools since 2025.
• MOE’s 2026 plans also add developmental AI-literacy milestones and strengthen cyber-wellness lessons on generated information and deepfakes.
• SLS AI tools can provide learning support, but generated responses can be inaccurate and do not determine a student’s G level.
• Responsible AI means verifying information, understanding a method, protecting private data and documenting the child’s own contribution.
• Parents need not buy an expensive coding kit or subscribe to a commercial AI product unless the receiving school specifically requires it.

Quick answer: Code for Fun develops computation and applied digital-making skills. In 2027 MOE is expanding AI literacy; ask what the school’s actual programme teaches and insist on independent understanding, safe data use and verified explanations.

Contents: Code for Fun and AI literacy in Full SBB

• 1. What Code for Fun actually is
• 2. What MOE announced for 2027
• 3. Code for Fun versus AI for Fun
• 4. How Full SBB fits with digital projects
• 5. The difference between ALP and Code for Fun
• 6. Why computation begins before writing code
• 7. What an algorithm is
• 8. Block-based programming
• 9. Text-based programming
• 10. Microcontrollers and real-world inputs
• 11. Building a useful prototype
• 12. Design thinking and empathy
• 13. A sustainability project example
• 14. A transport project example
• 15. What machine learning means at this age
• 16. Generative AI versus ordinary automation
• 17. Why prompt wording matters
• 18. How a generated answer can be wrong
• 19. Hallucinations and false references
• 20. AI in the Student Learning Space
• 21. Code for Fun and digital citizenship
• 22. Protecting children’s personal data
• 23. Bias in an AI system
• 24. Cyber wellness and deepfakes
• 25. Computer science versus using an app
• 26. What if a pupil has never coded before?
• 27. How to troubleshoot without guessing
• 28. Teachers and student ownership
• 29. How to assess a digital project
• 30. The relationship to school ALPs
• 31. How this connects with Mathematics
• 32. How this connects with English
• 33. How this connects with Science
• 34. How to choose further opportunities
• 35. What it may cost
• 36. Punggol families and the practical school week
• 37. A fictional project with errors
• 38. A four-week exploration plan

1. What Code for Fun actually is

Code for Fun is a programme jointly offered by MOE and IMDA to expose school students to computational thinking, coding, design and technology through applied learning. It is not the name of a national examination subject and does not automatically add another SEC paper to the student’s timetable.

A practical session may involve discussing a problem, testing a simple algorithm, using a sensor or creating a prototype. The point is learning how a process works, not completing a worksheet of mysterious code.

Try this: Ask the teenager to describe what they made, which problem it addresses and what they would change.

2. What MOE announced for 2027

MOE announced at the March 2026 Committee of Supply debate that the refreshed Code for Fun programme would strengthen AI skills and be available to all schools in 2027. It also set out clearer developmental milestones for AI literacy throughout the curriculum, co-curriculum and self-directed learning.

The announcement establishes national policy direction. It does not provide a universal classroom-by-classroom schedule, promise every school will purchase identical hardware or set a standard 2027 AI assessment paper. School communication remains essential.

Try this: Check the receiving school’s 2027 technology and applied-learning notice before buying equipment.

3. Code for Fun versus AI for Fun

Code for Fun is the broader programme that introduces computation, digital making and emerging technologies. IMDA’s description says AI for Fun modules became available to students in government and government-aided schools from 2025, with examples such as generative AI, smart robots and digital making using AI.

A family may see both names in school presentations and assume one replaces the other. They are related initiatives. The school decides the approved modules and sequence for its students.

Try this: Ask which learning objectives are covered rather than assuming the label guarantees a particular robotics workshop.

4. How Full SBB fits with digital projects

Full SBB allows students to study academic subjects at different G levels and learn together in mixed form classes. A digital project can draw on Mathematics, Science, language, art, social understanding and teamwork, even when students study these academic subjects at different levels.

The classroom challenge is to give each pupil an appropriate role and learning opportunity. A child strong at visual design may contribute a clear interface, while another reasons through an algorithm and a third asks whether the solution helps the people it targets.

Try this: Celebrate contributions through evidence of learning instead of assigning technological potential by Posting Group.

5. The difference between ALP and Code for Fun

An Applied Learning Programme (ALP) is a school’s distinctive approach to connecting academic knowledge with real-world problems. Code for Fun can complement an ALP through design, computing and prototyping. The programmes are not interchangeable: a school’s ALP may focus on sustainability, languages, business or other domains.

IMDA notes that Code for Fun can connect with ALP themes such as sustainability and transportation. A family should check what the school itself actually offers, especially when considering schools in Punggol.

Try this: Ask which project belongs to the national Code for Fun programme and which to the school’s own ALP.

6. Why computation begins before writing code

Computational thinking includes breaking a problem into manageable parts, recognising patterns, identifying a repeatable process and describing instructions precisely. It is useful in coding but also in household planning, experiments and everyday reasoning.

A teenager can practise sequencing and debugging without learning every feature of a programming language. This is particularly helpful for students who fear that coding means memorising hundreds of technical words.

Try this: Have the student explain an algorithm in ordinary English before translating it into code.

7. What an algorithm is

An algorithm is a set of steps for solving a class of problems. A recipe, timetable check or method for sorting items can illustrate the idea, although actual computer algorithms require precise representations and well-defined behaviour.

The child should notice the difference between an instruction that a person interprets flexibly and an instruction a computer follows exactly. Ambiguity can create errors even when the basic idea is sound.

Try this: Ask the teenager to give an instruction that fails because it is incomplete, then improve it.

8. Block-based programming

Block-based environments represent programming operations using visual blocks. They can help beginners build sequences, conditions, loops and events without needing to memorise complex syntax first. IMDA’s earlier Code for Fun framework allowed schools to choose this track.

A block-based activity is not automatically inferior to a text-based one. A sophisticated problem can still involve careful conditions, debugging and interaction with physical devices.

Try this: Ask what the blocks do and why the order matters instead of comparing the interface’s appearance.

9. Text-based programming

Text-based programming represents instructions through written syntax and language rules. It can be suitable when students are ready to manage variables, conditions, functions or hardware commands in a more explicit form. The school’s selected tools and expectations determine what is taught.

A teenager who types a correct-looking script may still be copying without understanding. The more important evidence is that they can predict the program’s behaviour, explain a mistake and improve a result.

Try this: Encourage independent explanation and testing before purchasing a commercial coding course.

10. Microcontrollers and real-world inputs

A microcontroller is a small computing device used in physical projects. It can read a sensor, process simple information and control an output such as a light, motor or buzzer. IMDA describes hands-on digital making through microcontrollers in its secondary programme.

Students can see an abstract instruction become a physical action. This creates useful opportunities to discuss measurement, electrical safety, design constraints and unreliable sensor readings. The precise hardware is chosen by the programme and school.

Try this: Ask how the prototype senses a change and what it does in response.

11. Building a useful prototype

A prototype is an early version of a solution used to test an idea. It might be made from cardboard, a simple circuit or a digital model. The value lies in discovering what works and what needs revision before investing heavily in a final design.

Parents can support prototyping by asking about the intended user and the problem. Buying polished materials or finishing the model on the child’s behalf can conceal the learning process.

Try this: Ask for the first unsuccessful version and what the pupil learned from it.

12. Design thinking and empathy

Design thinking begins by understanding a human problem before committing to a technical answer. Students may talk to potential users, observe constraints, generate options, make a prototype and seek feedback.

This helps prevent technology projects that look impressive but solve no real need. It also links computation with communication and responsibility, showing why a child’s English and social reasoning matter alongside technical skill.

Try this: Ask whose need the project addresses and what evidence confirms that need.

13. A sustainability project example

Imagine a fictional school challenge involving water use in a small garden. Pupils could investigate how a sensor reports moisture and how to decide when watering might be needed. This connects technology with environmental observation and scientific reasoning.

The students should not claim that a simple sensor automatically produces optimal irrigation. They need to examine accuracy, environmental effects and possible false readings. Real-world testing is part of learning, not a reason to reject the project.

Try this: Have the pupil distinguish sensor measurement, program rule and actual water-saving evidence.

14. A transport project example

A class could model how a crossing signal helps people move safely, or how changing bus arrival intervals affects waiting. Such a project can connect sequencing and conditions with real transport considerations in Singapore.

Pupils should not treat a classroom model as authority to alter public transport infrastructure. The purpose is to understand a design decision and its trade-offs.

Try this: Ask what the model leaves out compared with a real Punggol pedestrian crossing or bus service.

15. What machine learning means at this age

Machine learning refers to systems that learn patterns from data rather than having every possible response manually written as a fixed rule. A student might encounter a simple classification example and discover that the quality of the training data affects results.

This is not permission to use real classmates’ private images or personal records without authorisation. Demonstrations should use approved, appropriate materials and clear privacy safeguards.

Try this: Ask the child what the system learned from, how it was tested and where errors could occur.

16. Generative AI versus ordinary automation

A traditional programmed rule might turn on a light when a sensor crosses a threshold. Generative AI can produce new text, images or other content based on patterns learned during training. Its output may sound convincing even when incorrect.

Understanding the difference helps pupils avoid treating a generated paragraph as an official source. It also explains why learning fundamental programming and logic still matters in an AI-rich world.

Try this: Ask the child to distinguish an observed measurement from a generated prediction or explanation.

17. Why prompt wording matters

When pupils use an approved generative AI tool, the wording of a request can influence what the system produces. Clear context, a specific learning objective and a requirement to show reasoning may make an answer more useful, but they do not guarantee truth.

A child should not infer that a fluent response means the model has verified every claim. Prompting is a communication skill that needs verification and judgement.

Try this: Have the student compare two explanations and identify which claims need independent checking.

18. How a generated answer can be wrong

AI systems can produce invented citations, incorrect dates, faulty calculations or plausible explanations built on false premises. MOE’s SLS guidance warns that generated outputs are probabilistic and may be inaccurate.

A teenager needs a process for checking results: identify the claim, find a reliable source, test an example and ask the teacher when uncertainty remains. Schoolwork can become more accurate through this habit even without advanced technical knowledge.

Try this: Practise checking one surprising AI-generated statement against the original lesson or official source.

19. Hallucinations and false references

A fabricated reference is dangerous because it may look exactly like a trustworthy citation. Pupils might copy a made-up book title or website link into a project without checking whether the source exists.

Teach the child to open the source, verify author and date, and look for evidence that directly supports the claim. The ability to verify matters more than the number of references displayed.

Try this: Do not include any source in a school assignment that the student has not actually checked.

20. AI in the Student Learning Space

The Singapore Student Learning Space hosts approved learning activities and has introduced AI-enabled support features. Teachers determine how these tools fit learning objectives; pupils do not automatically receive unrestricted access to every external AI service.

Parents should separate features inside approved school systems from consumer tools with their own account requirements and terms. A student’s academic subject level does not itself authorise a personal AI subscription.

Try this: Use the school’s platform guidance and ask what assistance is permitted in each task.

21. Code for Fun and digital citizenship

Learning to code includes decisions about safety, fairness, consent and social impact. A pupil who creates a survey or simple app must consider the information it collects, how it will be protected and whether users understand its purpose.

Technical success is not enough if an activity exposes classmates’ private information or creates an unfair outcome. Teachers can use small projects to connect engineering with responsible decisions.

Try this: Ask what information the prototype needs and whether collecting less data would be safer.

22. Protecting children’s personal data

Student names, identity details, images, grades, health records and private messages should not be pasted into unfamiliar online tools without appropriate permission. A seemingly harmless classroom demonstration can create privacy risks if real data are used unnecessarily.

Parents can model good habits by using fictional examples and checking a service’s approved use before uploading work. School rules and applicable data-protection practices govern real student information.

Try this: Replace personal details with invented examples when testing a public digital tool.

23. Bias in an AI system

An AI model can perform unevenly when its data do not represent the people or conditions in which it is used. A speech tool, for example, may handle one accent better than another. Students can discuss whether a technological result is fair and reliable.

Children should learn to question the data and test across appropriate examples instead of assuming that a machine output is neutral. This habit connects technology with empathy and careful evidence-based thinking.

Try this: Ask who might be disadvantaged if a system makes a particular kind of error.

24. Cyber wellness and deepfakes

MOE’s 2026 announcements include updated Character and Citizenship Education lessons on checking generative AI information and recognising deepfakes. Students can encounter manipulated images, voice recordings or convincing online claims outside school.

A child who receives an alarming video should not automatically share it. It is better to pause, examine the source, seek context and report harmful or deceptive content through suitable channels.

Try this: Teach a simple routine: stop, verify, ask a trusted adult and avoid amplifying the deception.

25. Computer science versus using an app

Using a digital app and understanding how a program works are not the same skill. A pupil may quickly navigate a video editor but struggle to describe its logic, inputs, limits or privacy settings.

Code for Fun can help reveal the reasoning beneath an interface. Parents should not assume that a teenager who spends many hours on a screen has developed computational thinking automatically.

Try this: Ask the child to explain one digital process using inputs, decisions and outputs.

26. What if a pupil has never coded before?

The programme is intended to expose students to technology and build skills. Beginners can learn through clear examples, visual representations and guided practice. Prior paid coding lessons are not a universal prerequisite for participating in school activities.

Students learn at different rates. A child comfortable in Mathematics may still need help debugging physical hardware, while another who finds symbolic algebra challenging may excel at practical design.

Try this: Treat mistakes as information about the next concept to teach, not proof of low technological potential.

27. How to troubleshoot without guessing

Debugging means identifying why a system behaves differently from what was intended. A useful student checks one input or condition at a time, observes the result and records the cause. Randomly changing many lines of code often makes the problem harder.

The habit resembles scientific investigation: control variables, collect evidence and revise a hypothesis. It also applies to everyday schoolwork, especially mathematical reasoning and reading instructions carefully.

Try this: Ask which change produced the difference instead of immediately fixing the project for the child.

28. Teachers and student ownership

A teacher can model a coding technique and give feedback, but the pupil needs opportunities to make decisions. A parent can offer guidance while avoiding the temptation to produce a professional-looking project that the child cannot explain.

Authentic work includes planning notes, failed tests, corrections and reflections. The final artefact matters, but the learning process explains how the teenager acquired competence.

Try this: Keep a short record of the child’s own choices and what they learned from feedback.

29. How to assess a digital project

A project can be judged by problem definition, clarity of method, successful implementation, testing, communication and ethical awareness. The exact school rubric may differ and should be read before work starts.

An attractive prototype may fail to address the problem, while a modest one may demonstrate careful reasoning. Parents should resist ranking every project by how many electronic parts it uses.

Try this: Ask which evidence demonstrates that the proposed solution works for its intended user.

30. The relationship to school ALPs

Greendale Secondary’s published STEM ALP includes Design, Technology and Engineering experiences such as coding, electronics and designing responses to community challenges. Other schools emphasise sustainability, entrepreneurship or different applied domains.

These examples show that technology can be embedded in a broader learning environment. They do not guarantee the same modules at every campus. The child’s actual school determines the programme and selection opportunities.

Try this: Compare learning goals and school culture, not only the hardware featured in open-house photographs.

31. How this connects with Mathematics

Programming often uses sequences, variables, conditions and patterns. These ideas can reinforce mathematical reasoning when students understand the connection. For example, a sensor threshold can introduce comparison and inequalities in an accessible context.

It is important not to claim that programming experience automatically raises the student’s G level or examination grade. Formal subject-level changes require school assessment and readiness.

Try this: Use a project to motivate clear mathematical explanations, not as a substitute for learning the syllabus.

32. How this connects with English

Designing a useful tool requires explaining a problem, writing understandable instructions, presenting evidence and listening to feedback. These are language skills. A student who communicates a prototype clearly can make the work more useful to others.

AI-generated prose can be tempting, but a student should remain capable of explaining what was built and why. Schools may have academic-integrity rules about digital assistance.

Try this: Ask the teenager to present the project in a short explanation without copying a generated script.

33. How this connects with Science

Sensors, motors and simple models create opportunities to consider measurements, energy, environmental systems and fair tests. Pupils can learn that a prototype’s readings vary and that a result must be repeated to establish confidence.

A model is not automatically a full scientific experiment. Students should describe uncertainty, limitations and the evidence needed before making a strong claim.

Try this: Ask how the pupil would test the same idea again with one controlled change.

34. How to choose further opportunities

A child who enjoys a short Code for Fun activity may want to explore a school club, an ALP extension or an approved competition. Those opportunities depend on interest, age, school provision and any selection rules.

Families do not have to purchase every available coding programme immediately. A small, sustained project and careful feedback can be more educational than rapidly switching between subscriptions and gadgets.

Try this: Begin with the child’s genuine interest and one small next challenge.

35. What it may cost

IMDA’s published secondary programme describes programme funding for participating government and government-aided schools, including vendor-supported training hardware. That does not imply every optional external course or personal robotics kit is free.

Parents should follow school-issued requirements for PLDs, software and project equipment. If a genuine financial barrier arises, ask the school privately about available support before purchasing alternative hardware.

Try this: Confirm the actual mandatory purchases and school-provided resources before paying.

36. Punggol families and the practical school week

A teenager may have G-level academic lessons, shared form-class experiences, CCA, HBL and digital projects in the same week. A technology activity should enhance curiosity without crowding out sleep, friends and physical movement.

The real commute from Punggol and any after-school programme also matter. Students need time to test ideas calmly, and an overloaded schedule can reduce the learning value of a good opportunity.

Try this: Discuss the whole week’s commitments before accepting an optional enrichment activity.

37. A fictional project with errors

Imagine fictional student Kai building a simple light-responsive model. It works indoors but fails near a bright window. His first instinct is to copy a different program from the internet.

A teacher helps him identify a changing sensor reading and adjust the testing process. Kai learns more by understanding the failure than he would by submitting a borrowed script that happens to run.

Try this: Ask the pupil which observation changed their explanation of the problem.

38. A four-week exploration plan

Week one: learn what inputs and outputs are and examine a simple example. Week two: create a sequence or block program that performs a small task. Week three: test and debug it using evidence. Week four: explain the result, its limits and any privacy or safety concerns.

This is a home exploration framework, not a promise about the school’s 2027 Code for Fun schedule. If the student is working on an actual school assignment, follow that teacher’s objectives, tool rules and deadline first.

Try this: Keep a brief reflective record of what the child built, tested and understood.

Parent and student checklist

• Confirm the school’s actual 2027 programme and module choice
• Understand Code for Fun versus AI for Fun and SLS
• Check prescribed hardware, accounts and fees before buying anything
• Ask which problem a project solves and who its users are
• Explain how inputs, instructions and outputs relate
• Test a prototype and record why a result changed
• Verify AI-generated claims instead of trusting fluent writing
• Protect students’ personal data and avoid uploading private school information
• Observe school academic-integrity rules for assisted work
• Review whether enrichment fits the student’s ordinary school and CCA commitments

Frequently asked questions

Is Code for Fun compulsory in every secondary school in 2027?

MOE said the updated programme would be available to all schools. Exact implementation and pupil participation should be checked with the school.

Will all pupils learn Python?

No specific programming language is guaranteed nationally. The earlier framework included block- and text-based tracks.

Is Code for Fun the same as SLS?

No. Code for Fun is a technology-learning programme; SLS is a digital learning platform.

Does G1 prevent a child from learning coding?

No. G levels describe specific academic subject demands, not a general judgement about technological creativity.

Will every child receive a robot?

No. Hardware and activities depend on the school’s approved programme.

What is AI for Fun?

Modules exposing students to AI concepts, generative AI, robotics and responsible technology use within the national programme.

Can AI-generated answers be wrong?

Yes. Students should verify claims with reliable evidence and teacher guidance.

Should parents buy a coding kit before school begins?

Not unless the actual school specifies that it is needed.

Can pupils paste private data into an AI tool?

They should follow school privacy rules and avoid disclosing classmates’ personal information to unapproved services.

Does doing an AI project guarantee a future computing course?

No. Later choices depend on subjects, results, interest and admissions criteria.

Where do we verify the plans?

Use the MOE 2026 announcements, IMDA Code for Fun guidance and the allocated school’s programme notice.

Official sources and next reading

Source checks: MOE Committee of Supply 2026: AI literacy and Code for Fun in 2027; IMDA Code for Fun; IMDA secondary programme; SLS Guidance on Generative AI, updated September 2026; Greendale Secondary ALP. Related guide: Full SBB HBL, SLS and personal learning devices.

Editorial review: 6 November 2026. Exact 2027 school modules, equipment and learning schedules are not universal. All fictional project examples are educational illustrations, not accounts of named pupils.