After PSLE results, parents preparing a child for Secondary 1 in 2027 should know that artificial intelligence will increasingly be part of digital literacy and school learning. MOE’s 2026 education announcements say AI literacy is being strengthened through curriculum, co-curricular activities and self-directed resources, and the updated Code for Fun programme will be available to all schools in 2027. Cyber Wellness education also includes checking AI-generated information and recognising deepfakes. This does not mean every Secondary 1 child must buy an expensive AI subscription or use a chatbot to complete homework.
The most useful home preparation is to establish a clear rule: your child should learn the underlying ideas, attempt assignments independently, follow the teacher’s AI-use instructions and check any AI-generated explanation against reliable sources. MOE’s AI in Education Framework highlights Agency, Inclusivity, Fairness and Safety, while the Student Learning Space provides official guidance for teachers and parents on responsible use of generative AI. AI can explain, quiz or help students notice an error, but it can also invent facts, misinterpret a question or produce polished work that hides the student’s lack of understanding. The family should encourage critical thinking and academic honesty rather than automatically using or banning every tool.
- Know the 2027 change: MOE is updating Code for Fun to strengthen student AI skills, with programme availability across schools.
- Understand the distinction: Code for Fun, optional AI for Fun modules, school AI literacy activities and personal chatbot use are not identical programmes.
- Read school rules: follow the actual teacher’s directions for when AI help is permitted, what must be original and how assistance should be acknowledged.
- Start with understanding: explain Mathematics methods, Science ideas and written arguments before asking an AI tool for answers.
- Verify AI output: check calculations, quotations, dates, source links and claims against trustworthy references.
- Protect privacy: do not upload examination slips, student identity numbers, medical reports or classmates’ images into an unapproved tool.
- Recognise synthetic media: pupils should learn to question deepfakes, manipulated pictures and invented audio.
- Use age-appropriate access: check each service’s terms, the school’s approved platforms and family supervision requirements.
Definition Lock: What AI Literacy Means for a Secondary Student
AI literacy means having enough understanding and judgement to use, interpret and question AI-enabled technologies responsibly. It includes knowing what a tool can do, where its output may be unreliable, how data and bias affect systems and what the user remains responsible for deciding.
A Secondary 1 child does not need to become an AI researcher or understand the mathematics of every neural network. They do need to recognise that a confident answer can be wrong and that an automated system is not a substitute for thinking or ethical responsibility.
The educational goal is broader than typing effective prompts. Students should understand why the tool behaves as it does at a basic level, what information they should not share and when a task requires their own independent work.
Parents can support these habits through ordinary conversations about school assignments, online images and information sources without buying specialist AI courses before the child has entered secondary school.
What MOE Announced for 2027
MOE’s 2026 Committee of Supply announcements describe AI literacy being incorporated into curriculum, co-curricular learning and self-directed resources, with defined developmental milestones through the education journey.
The ministry also said the Code for Fun programme would be updated to strengthen students’ AI capabilities and made available to all schools in 2027. This is an availability commitment at school level, not a claim that every pupil must take the same AI elective on the first day of Secondary 1.
The Cyber Wellness curriculum has been updated to include validating information generated by AI and identifying deepfakes. This connects technological understanding with how students protect themselves and others online.
Families should use the receiving school’s current 2027 notices for actual programme timing. A national rollout announcement does not set an identical workshop date for every campus.
Code for Fun Is an Educational Programme
Code for Fun introduces age-appropriate coding and computational thinking through structured school experiences. With the 2027 updates, MOE intends to strengthen AI-related skills within the programme.
Coding and computational thinking help students break problems into steps, identify patterns and understand how instructions produce results. AI learning adds questions about training data, automated predictions and system behaviour.
The programme’s exact level and schedule can differ by school, so parents should ask whether the child will experience a module in Secondary 1, Secondary 2 or another appropriate period.
Do not confuse access to Code for Fun with automatic acceptance into a competitive school Computing CCA or upper-secondary Computing subject. Those opportunities can involve separate requirements.
AI for Fun Is a Separate IMDA Opportunity
IMDA’s AI for Fun initiative has offered a ten-hour module for secondary schools from 2025. Its published description includes training data quality, bias, large language models, custom chatbots, prompts, social consequences and application to real-world themes.
Under the existing programme arrangement, eligible government and government-aided secondary schools apply, and students are expected to complete Code for Fun or a comparable programme before taking AI for Fun.
That does not mean every Secondary 1 student is automatically enrolled in AI for Fun as soon as the national Code for Fun update becomes available. The two programmes have different administrative arrangements.
Ask the receiving school what it actually offers and when. A meaningful AI education can include lessons and responsible-use practices even when the school does not run the same optional module as another campus.
Four Principles in MOE’s AI Education Framework
The official AI in Education Framework sets out four guiding principles: Agency, Inclusivity, Fairness and Safety. These help educators decide how and when AI should support learning.
Agency means students and teachers retain meaningful responsibility for judgement and decisions. Inclusivity concerns appropriate access and participation for learners with different needs. Fairness includes attention to inequitable or biased outcomes. Safety includes protecting users and using systems responsibly.
A parent does not need to memorise the four words as an examination list. Their value comes from applying them: Who is making the decision? Can everyone participate? Could an output be biased? What information or harm needs protection?
These questions apply whether a child is using an SLS learning tool, reading AI-generated material or considering a commercial chatbot at home.
Agency: The Child Still Does the Thinking
Agency is the ability to make meaningful choices, understand their effects and remain responsible for one’s work. An AI system can propose a plan, but the student must decide whether it is appropriate and explain why.
A child who copies an AI-generated Mathematics solution without understanding it may submit correct-looking work while learning little. A student who attempts the problem, then compares the explanation and identifies an error, is using the tool more constructively where permitted.
Parents can ask, ‘What did you decide yourself?’ and ‘Which part did the tool help you understand?’ This makes the student think about the boundary between assistance and replacement.
The same principle applies to writing, research and presentations. The pupil’s own judgement is the central educational outcome.
Inclusivity: Different Learners Have Different Needs
AI-enabled tools may support access to information through features such as text-to-speech, language explanations or other approved learning aids. These possibilities can be helpful for students with differing strengths and needs.
However, not every tool works equally well for every learner, and an accessible-looking feature can introduce new errors or frustrations. The actual school’s authorised arrangements and any professional recommendations should guide use where relevant.
A family should not assume that purchasing an expensive AI service is necessary to ensure the child receives adequate academic support. Schools can provide resources and guidance through approved processes.
Inclusive education means the right help at the right time, not giving every student an identical technological solution regardless of need.
Fairness: AI Can Produce Unequal or Biased Results
AI systems learn patterns from data and can produce outputs that reflect stereotypes, omissions or other biases. A response may favour particular languages, social contexts or interpretations without explaining the limitation.
Students should learn to question sweeping statements about groups and verify claims about people. A model’s confident style does not make its conclusions fair or accurate.
A student might ask an AI tool to describe a profession and receive an answer implying it is suitable only for one gender. The family can discuss why that is an unsupported assumption and how real skills and interests matter.
Responsible use includes noticing these problems rather than repeating them in schoolwork or group chats.
Safety: Privacy and Wellbeing Matter
AI tools may process information that users enter. Children should not upload classmates’ names, private messages, identification numbers, results slips, photographs or medical information to an unapproved system.
Parents should also consider the emotional effect of asking a chatbot for answers about personal relationships or serious health concerns. A tool can make mistakes and should not replace trusted adults or qualified professionals.
Teach the child to stop and ask for help when a generated response seems unsafe, frightening or inappropriate. Age restrictions and school policies should be respected.
Safe use is more than enabling a filter. It includes judgement about data, trusted people and the purpose of the interaction.
What Generative AI Does
Generative AI can produce new text, images, audio and other outputs in response to prompts. It works by using patterns learned during training, rather than reading a teacher’s mind or knowing whether every generated claim is true.
A large language model can produce fluent sentences that seem authoritative. It may also confuse two historical events, invent a reference or make an algebraic error while maintaining a convincing tone.
Explain this simply: the tool predicts and generates patterns of language; the student must check whether the result matches the real question and reliable evidence.
This basic understanding reduces the risk of treating the chatbot as an infallible authority.
AI Can Sound Right and Still Be Wrong
One of the hardest parts of evaluating generative AI is that incorrect answers can be expressed with excellent grammar. A student may assume that a polished paragraph or sophisticated equation explanation must be correct.
Teach the child to verify specific claims. If the response names an official MOE rule, open the current MOE page. If it gives a historical date, consult a trustworthy source. If it performs a calculation, check the result independently.
Do not rely on how confident or friendly the tool sounds. Fluency and factual accuracy are different qualities.
The child should also understand that AI may produce a different answer to the same question on another occasion. Apparent consistency is not proof of truth.
What Is an AI Hallucination?
An AI hallucination is generated information presented as though it were factual or supported, even though it is false or not grounded in reliable evidence. It can include invented books, quotations, sources or numerical claims.
A Secondary 1 pupil writing a History project could receive a fabricated speech attributed to a real historical figure. If the student copies it without checking, the final work becomes unreliable.
Parents can model the habit of seeking the original source. If a quotation cannot be found in a trustworthy document, it should not be used as confirmed evidence.
Do not turn the term into a joke suggesting every AI output is meaningless. It is a reason to verify important material, especially when a claim affects someone else.
Why Prompting Is Not the Whole Skill
A clear prompt helps a tool understand what kind of response is requested. It can specify the learning topic, reading level and desired explanation format, subject to the service’s permitted use.
But an excellent prompt cannot guarantee a correct answer. The user still needs domain knowledge, source checking and critical judgement.
Teach the child to ask focused questions, such as ‘Explain why the same operation must be applied to both sides of an equation,’ rather than ‘Do my entire Mathematics homework.’
This encourages the student to learn a concept instead of merely obtain a finished product.
An AI Tutor Is Not a Replacement for the Teacher
AI tools can provide additional explanations or practice, but they may not know the school’s current syllabus level, teacher’s method or assessment instructions.
A student taking G1 Mathematics can be confused by an answer using advanced G3 techniques. A Secondary 1 English learner may receive writing advice unsuitable for the genre assigned.
When a tool conflicts with the teacher’s instruction, the child should seek clarification through school channels rather than assume the AI is always right.
The most productive approach is to treat authorised tools as supplementary aids while keeping the actual curriculum and professional teaching central.
What Is Allowed in Homework?
There is no safe assumption that every school teacher permits identical AI assistance for every assignment. A task may ask for independent writing, supervised research or appropriate use of digital tools with disclosure.
The pupil should read the actual instructions and ask if uncertain. For one task, generating discussion questions might be acceptable; for another, submitting AI-written paragraphs as original work may be prohibited.
Parents should avoid declaring that because AI is available, every use must be legitimate. Availability is not permission to bypass assessment expectations.
A child who understands the difference between help and misrepresentation develops academic integrity that will matter throughout education.
Academic Honesty Is About Who Did the Work
A student who submits an AI-generated essay while claiming it is entirely their own has misrepresented authorship if the task requires independent writing. The problem is not that technology exists; it is that the work does not accurately demonstrate the learner’s understanding.
Help the child distinguish brainstorming, feedback, proofreading and generating a complete answer. The teacher’s rules determine which forms of assistance are allowed and how they should be acknowledged.
A parent should not secretly use AI to rewrite every assignment into advanced language. That can hide the child’s real ability and make teacher feedback less useful.
The goal of schoolwork is learning and an accurate demonstration of knowledge, not merely producing a polished page.
The Citation Problem
AI tools sometimes generate references that look like genuine articles but have incorrect titles, authors or links. A student could accidentally cite a source that does not exist.
Teach the child to open each source independently, check its author or publisher, confirm the date and verify that it actually supports the claim.
Do not copy an automatically generated bibliography without reading the original material. A legitimate source may also be outdated or irrelevant to the 2027 school context.
This is especially important for Humanities and Science projects, where claims should follow evidence rather than decorate a paragraph with impressive-looking citations.
Mathematics: Use AI to Understand, Not Copy
If an AI tool produces a solution, ask the student to explain every step and check the result by substitution or another valid method.
For an equation such as 3x + 5 = 20, the correct solution is x = 5. The child should understand subtracting 5 and dividing by 3, not merely read the final answer.
If the AI proposes an incorrect move, ask which mathematical rule it violates. This is an opportunity to build reasoning, provided the teacher permits the use.
Do not outsource every homework problem. A child needs independent practice to know whether they can solve unfamiliar questions without the tool.
Science: Ask About Evidence and Conditions
Science often involves data, variables and explanations. An AI system may describe an experiment confidently while missing a relevant control or making an unsafe suggestion.
Students should use the actual school syllabus and teacher-approved materials for practical tasks. Do not perform hazardous chemistry or electrical experiments because a chatbot produced a convincing procedure.
For written reasoning, ask which observation supports the explanation and whether the AI has confused correlation with cause.
The student’s ability to connect evidence with scientific ideas remains more important than generating a long explanation quickly.
English: Preserve the Student’s Voice
An AI writing tool can transform a simple paragraph into elaborate prose. That may look impressive, but it can remove the child’s original ideas and conceal the skills they need to develop.
If the teacher allows feedback assistance, a more educational use might involve asking for one explanation of a grammar error, then having the student revise their own work.
A student should still decide the main argument, select evidence, organise paragraphs and explain the final text.
Do not teach that good writing always means the most complicated vocabulary. Clarity, purpose and truthful authorship matter more.
Humanities: Check Dates and Context
A child researching Singapore history or a world event may use AI to generate an overview. The tool could confuse chronology, omit important context or fabricate a quotation.
Teach the student to verify dates and names against reliable historical sources and to distinguish original evidence from commentary.
If two sources disagree, ask which evidence supports each interpretation rather than automatically choose whichever explanation sounds more confident.
Humanities is particularly valuable for teaching careful source evaluation, a skill students will need when AI-generated information becomes widespread.
Deepfakes and Synthetic Images
A deepfake is manipulated or generated audio, video or imagery that can falsely depict a person saying or doing something. Students may encounter these materials through social media or group chats.
MOE’s updated Cyber Wellness education includes recognising deepfakes and validating AI-generated information. Parents can reinforce the skill by explaining that realistic appearance does not prove authenticity.
Do not encourage children to generate fake images of classmates or teachers as pranks. Such content can create privacy, dignity and bullying harms.
When a harmful fake appears, the child should seek a trusted adult and use appropriate reporting channels rather than redistribute it widely.
Protect the Child’s Personal Data
PSLE results slips, school login credentials, identification numbers, private family addresses and medical reports contain sensitive information. There is rarely a legitimate reason to upload them to an unapproved public chatbot for basic study assistance.
Teach the child to replace personal details with fictional examples when asking general questions. A Mathematics problem does not require the student’s real name or home address.
School learning platforms may have authorised data arrangements, but parents should still follow the institution’s guidance.
Privacy protection is a habit that should begin before a child becomes accustomed to sharing information with every app offering homework help.
Cyberbullying and AI-Generated Content
AI can make it easier to produce persuasive false messages, humiliating images or impersonations. A child should understand that creating or forwarding such material can hurt classmates and lead to serious consequences.
If the pupil receives harmful content, advise them not to retaliate and to seek a suitable adult. Evidence should be handled securely rather than circulated to embarrass the affected person further.
The school’s own bullying reporting and support policies continue to apply where student relationships are involved.
Responsible AI literacy includes considering what a generated output does to other people, not only whether it is technically possible to create.
Different Student Needs and Accessibility
Some pupils may benefit from school-approved text-to-speech, language scaffolds or other digital tools. These can improve access when used appropriately.
However, AI explanations can also contain errors that particularly affect students who depend on them. A child may need teacher guidance to recognise unreliable suggestions.
For pupils with special educational needs, share relevant functional information through authorised school channels and follow professional recommendations where applicable.
The aim is genuine learning access and agency, not a requirement that every student use one identical AI tool.
Should Parents Buy an AI Subscription?
Not automatically. MOE provides resources through the school system and the Student Learning Space, while the school determines approved tools and curricular expectations.
A paid subscription does not guarantee more accurate explanations or better examination grades. It also may not meet the service’s age requirements or the school’s data and academic-integrity rules.
Before paying, ask what problem the subscription would solve and whether the child already has appropriate learning support through school.
A good educational purchase follows a defined need, not marketing pressure after PSLE results.
Age Restrictions Matter
Some AI services have minimum-age requirements or require parent or guardian permission for younger teenagers. These terms can change, so check the actual provider’s current conditions before creating an account.
Do not misstate a child’s age or encourage them to bypass platform restrictions just to access a homework tool.
School-approved environments may use different access controls and supervision, and parents should follow those instructions.
The child’s digital wellbeing and data privacy are more important than the urgency of trying a fashionable application.
The Student Learning Space Is a Different Setting
MOE’s Student Learning Space is an official learning environment that can provide curated content and school-managed AI-supported experiences.
A teacher-designed learning activity in SLS is different from an unapproved consumer chatbot whose data handling and age restrictions may differ. Students should not assume that a tool recommended online automatically has the same school endorsement.
Use the authorised account and follow teacher instructions. If an SLS activity includes AI support, the purpose is still to improve learning and independent judgement.
Parents can consult the official SLS guidance on generative AI to understand the intended safeguards and limitations.
How to Test Whether AI Help Was Useful
After a pupil consults a permitted AI explanation, close the tool and ask them to explain the concept in their own words. Can they solve a similar Mathematics question? Can they cite real evidence for a History claim? Can they revise an English paragraph independently?
If the answer is no, the AI may have completed the task without creating understanding. Another teaching explanation or simpler practice may be needed.
If the pupil can apply the idea independently, the interaction may have supported learning.
This is a practical way to distinguish genuine progress from the temporary appearance of competence produced by copying a polished response.
A Simple Five-Step Verification Habit
Before using an AI answer, the student can check five things: whether it answered the actual question, whether the facts have reliable sources, whether calculations are correct, whether the language is appropriate to the subject level and whether the use complies with the teacher’s rules.
For an important factual claim, opening an official source may be necessary. For a numerical result, substitution or estimation can provide an independent check.
The child should also ask whether any private information was unnecessarily shared. That makes safety part of verification rather than an afterthought.
A short, consistent routine is easier to apply than a lengthy list of technical terms a twelve-year-old will never remember.
A Worked Example: An Incorrect Algebra Solution
Imagine a student asks an AI tool to solve 2(x + 4) = 18 and receives x = 7. The answer is wrong because substituting 7 gives 22, not 18.
A student who checks the equation can notice the error and solve correctly: expanding gives 2x + 8 = 18, so 2x = 10 and x = 5.
The educational point is not that AI always fails. It is that a fluent solution should be tested against the mathematics before being trusted.
A child who performs the check independently has learned something useful even if the tool’s explanation was initially misleading.
A Worked Example: An Invented Historical Quote
A Secondary 1 student researches a historical figure and asks an AI tool for a dramatic quotation. The tool provides a statement and an impressive-looking source title.
The student cannot find the quotation in the original source. They should not include it as a verified historical fact simply because the wording sounds plausible.
The family can help the pupil find an authentic document, use an accurately paraphrased idea or state that the proposed quotation could not be verified.
This illustrates how AI literacy and Humanities source evaluation reinforce one another.
A Worked Example: AI Rewrites a Whole Essay
A pupil drafts a short argument, then asks AI to produce a much longer and more sophisticated version. The final essay contains vocabulary the pupil cannot explain.
If the teacher asked for independent writing, submitting that rewrite as the student’s own work may breach the task’s expectations. Even if the teacher permitted limited feedback, the child has lost ownership of the ideas.
A better use, where allowed, is to ask for two questions about organisation and then revise the original text personally.
The goal is a student who can write and explain, not a document that looks impressive while hiding a learning gap.
A Worked Example: A Deepfake in a Class Chat
A classmate circulates an AI-generated audio clip making a student appear to insult somebody. Others assume the voice proves the accusation.
A responsible pupil should avoid forwarding it, consider the possibility of manipulation and seek adult guidance if harm or bullying is occurring.
Parents should not publish the clip widely to ask internet strangers whether it is fake, especially when it identifies children.
The example shows why MOE emphasises deepfake recognition as part of Cyber Wellness rather than treating AI only as a new method of answering questions.
A Worked Example: An AI Tool Helps With Vocabulary
A student encounters the word ‘inference’ in a passage and wants a simpler explanation. With an approved tool, they ask for a definition and an example, then compare it with the teacher’s material.
The pupil creates a new example independently and uses the term correctly in a discussion. This is closer to genuine learning than asking the tool to complete the entire comprehension exercise.
The parent can reinforce the habit by asking the child to explain what they learned without repeating the tool’s exact sentence.
Good AI help makes the student more capable after the screen is closed.
How to Discuss AI Without Fear
Some parents believe AI will make every child intellectually lazy. Others believe every pupil must use it for every assignment to avoid falling behind. Both extremes can distort educational decisions.
AI is a technology with strengths, limitations and risks. The educational response should focus on understanding, appropriate use, teacher guidance and critical thinking.
A child does not need a fearful lecture every time a chatbot appears. They need clear examples of when it is helpful, when it may be wrong and when it is inappropriate.
A calm family approach is more likely to encourage honest questions and disclosure of mistakes than a mixture of panic and unrealistic promises.
A Holiday AI Literacy Conversation
During December, ask the child to think about how they verify an online claim. Choose a safe, non-sensitive factual question and compare an AI response with an official or trustworthy reference.
Discuss whether the tool cited a real source and which details needed checking. Do not ask the child to generate personal data about classmates or images of real people without permission.
The exercise can be brief. Its purpose is to develop healthy scepticism and a verification habit, not to train the child to become an expert prompt engineer before January.
The actual school will provide curriculum opportunities and rules appropriate to its students.
The First Month of Secondary 1
Read the school’s digital learning and AI-use guidance. Some classes may use approved AI-supported activities while others require independent work without external assistance.
Help the child record assignments and know when a question should be asked of the teacher. Do not assume that a commercial tool can interpret the exact school assessment rules.
Review whether the child is checking sources, retaining privacy and understanding the work they submit.
AI literacy develops through repeated responsible choices, not a single workshop or certificate.
The Parent’s Responsibility
Parents can model good habits by admitting uncertainty, checking a source and correcting a mistaken claim. If an AI tool makes an error, show the child how you identified it rather than simply dismissing or defending the technology.
Avoid uploading private school records to unapproved services in the name of seeking help. The family should demonstrate the privacy boundaries it wants the child to respect.
Encourage academic honesty: using a tool where permitted, acknowledging relevant help and retaining responsibility for original thinking.
The child’s future learning will benefit more from these habits than from owning the newest technology.
Frequently Asked Questions
Is AI literacy part of secondary school from 2027?
MOE announced wider integration of AI literacy and an updated Code for Fun programme available to all schools in 2027. Actual learning experiences follow school implementation.
Is Code for Fun the same as AI for Fun?
No. They are related but distinct programmes. IMDA describes AI for Fun as an additional module with its own school application and prerequisite arrangements.
Must parents buy ChatGPT or another paid AI subscription?
No general MOE requirement mandates a commercial AI subscription for every Secondary 1 pupil.
Can AI complete a child’s English essay?
A tool may technically generate text, but submitting it as independent work can violate the teacher’s instructions and conceal the child’s learning.
Is every AI-generated statement reliable?
No. AI can generate false facts, invented sources or calculation errors. Students must verify important information.
What are the four MOE AI education principles?
Agency, Inclusivity, Fairness and Safety.
Is using AI for clarification always permitted?
Follow the task and school’s current instructions. The rules can differ between assignments and tools.
Can a child upload their PSLE result slip to an AI chatbot?
Avoid uploading private student identification and examination documents to unapproved services. Seek official school assistance for administrative questions.
What is an AI hallucination?
Information generated as if factual even though it is unsupported or false, such as an invented quotation or source.
What is a deepfake?
Manipulated or generated media that can falsely depict a person saying or doing something.
How can parents tell if AI helped the child learn?
Ask whether the student can explain and apply the idea independently after closing the tool.
Do schools provide official AI guidance?
MOE’s SLS information site publishes responsible-use resources, and individual schools set relevant instructions for their learning activities.
Let AI Support Learning Without Replacing the Learner
A child entering Secondary 1 in 2027 will encounter an education system placing greater emphasis on AI literacy, digital judgement and safe use. The central challenge is not obtaining access to the most powerful tool. It is knowing when a tool is suitable and retaining responsibility for thinking.
Use MOE’s official guidance, respect school rules, check evidence and protect privacy. A student who learns to ask good questions, notice errors and explain independently will gain more from AI than one who merely copies its answers.
Official resources: MOE 2026 AI literacy announcements · MOE SLS guidance on generative AI · MOE AI in Education Framework · IMDA AI for Fun for secondary schools · eduKateSG PLD and SLS guide · eduKateSG HBL and independent learning · eduKateSG independent study habits.