Artificial intelligence is becoming an increasingly familiar part of children’s learning and digital environment. From explaining difficult concepts and generating ideas to supporting creative projects and exploring new questions, AI tools are opening up new possibilities for young learners.

At the same time, its growing presence raises important questions for parents. How should families introduce AI? How can children use it responsibly without becoming overly dependent on it? Should AI be used for schoolwork? And perhaps most importantly, how can parents ensure that technology supports children’s own thinking rather than replacing it?

Both Grace Chua, Head of Communications, Southeast Asia at OpenAI, and Sher-Li Torrey, Founder of Mums@Work, emphasise the importance of helping children develop good judgement around AI. Their perspectives suggest that the conversation for families is not only about whether children should use AI, but also about how they can use it thoughtfully and responsibly including knowing when to question and verify information, protect their privacy, think for themselves and seek human guidance.

Healthy AI habits also begin at home. Both experts highlight the importance of age-appropriate guidance, open conversations and parents modelling responsible technology use, while also recognising the importance of healthy offline experiences.

Interviews with Grace Chua and Sher-Li Torrey

With these questions becoming increasingly relevant to families, our team spoke with Grace Chua, Head of Communications, Southeast Asia at OpenAI, and Sher-Li Torrey, Founder of Mums@Work, about helping children develop healthy and responsible AI habits.

Their interviews explore the opportunities AI presents for learning, the importance of independent thinking, academic integrity, privacy, creativity, resilience and the role parents can play in preparing children for an AI-enabled future.

Interview with Grace Chua, Head of Communications, Southeast Asia at OpenAI 

1. AI is becoming an increasingly common part of children’s education and everyday lives. From your perspective, what are some of the biggest opportunities AI presents for young learners, and what are the most common misconceptions parents should be aware of?

Grace Chua: Where I see the biggest opportunity is in making learning more personalised, creative and accessible. I see that both through my work at OpenAI and, increasingly, through the lens of a parent thinking about the kind of world my son will grow up in. We’re already seeing a strong connection between AI and learning in Singapore. Our data shows that around 43% of ChatGPT usage among 18- to 24-year-olds is related to learning and education.

I often hear from colleagues and friends who use AI together with their older children that it can be really helpful to grasp a difficult concept from school or asking for a simpler explanation, a different example or another way of approaching the problem. That’s where I think AI can be particularly powerful. A child can ask a question the moment it occurs to them, or turn an idea into something they can test and build. Used thoughtfully and with guidance, AI can shorten the distance between curiosity and exploration. It can help children move from simply consuming technology to using it to learn and create.

One common misconception I see is that parents feel they have to choose between embracing AI without limits and avoiding it altogether. I do not think either extreme is especially helpful. AI can be a valuable learning companion, but, like any technology or tools, it’s important to know its limitations. While models today are very advanced, there are still instances where responses can be inaccurate, incomplete or sound more confident than they should. Learning to question and evaluate what it produces is therefore an important part of learning to use it responsibly.

For me, the question is not whether our children will encounter AI because they will in today’s world. As a parent, I think much more about how we help them encounter it well with curiosity, appropriate boundaries and critical thinking, and with the confidence to know that they do not have to accept everything a technology tells them.

That is also part of the thinking behind ChatGPT for Teens, an experience that we launched on 18 August. If our system estimates someone is under 18 or they state their age is between 13 and 17, they are automatically placed into ChatGPT for Teens. It’s designed to help teens learn, think critically, deepen their understanding, and use AI with confidence. It provides stronger built-in safety protections for teens, including features to promote healthy use and additional controls for parents.

It also includes practical features to support healthy learning habits. Quizzes can help teens test their understanding; responsible-homework reminders can redirect apparent shortcuts towards Study Mode; and Study Hours can make guided learning the default at times chosen by teens or their parents. These work alongside Quiet Hours and more frequent break reminders. 

2. As more families begin exploring AI tools at home, what are some healthy habits parents can establish early to help children use AI responsibly, thoughtfully and creatively, rather than becoming overly dependent on it?

Grace Chua: For me, it starts with open conversations at home. I don’t think conversations about AI or technology more broadly should be one-time talks. I know there is rarely one perfect moment when we as parents sit down and cover everything. Rather, these conversations happen over time, in small moments, just like the ones we already have with our children about friendships, kindness, school, online behaviour, money and all the other things they need help navigating.

These conversations often work best when they begin with openness rather than rules. That is also why a recent MOE Singapore post resonated with me. It suggested that when a child turns to an AI chatbot, our first response should not automatically be to reach for parental controls. There may be something underneath that behaviour whether it’s a need to be heard, understood or helped without judgement.

I think it’s helpful to ask, what kind of support is right for my child now? I find that a much more useful question than trying to come up with one fixed rule for every family. Because the level of guidance and support will evolve with every stage of our child’s development. For example, a younger child may need an adult beside them, while an older student may be ready for more independence and responsibility. It is not about giving every child the same tool at the same age, but about matching access, guidance and responsibility to where they are.

Three habits I find useful are (1) be clear about when AI is appropriate (2) keep some use visible and shared, and (3) ask children to bring something of their own before they ask the tool for help.

With younger children, that might mean exploring together in a common space, answering a question, inventing a story or planning a screen-free activity, and keeping the rules very simple. For example, do not share personal information, technology does not always know best, and tell an adult if something feels confusing or uncomfortable.

As they enter teenage years, the conversation can become more nuanced. What counts as help, and what becomes substitution? When should you trust your own voice? How do you protect your privacy, acknowledge AI use and recognise when a personal or emotional question needs a human being? The boundaries should adapt as they grow, but the underlying message stays the same. AI can support your thinking; it should not replace it.

Technology has a role to play too. With ChatGPT for Teens, which we just announced, those protections become part of the default experience for users we identify as under 18. That includes stronger safeguards in higher-risk areas, reminders that caution teens against uploading sensitive images, and clearer product behaviour that reinforces healthy boundaries and encourages offline support when it matters.

Those protections matter, but they are a starting point, not the whole answer. No setting can replace the conversations families need to have. For personal, emotional or high-stakes questions, children should know that a parent, teacher, healthcare professional or another trusted adult who understands their circumstances belongs in the conversation.

3. How can parents help children use AI as a learning tool while developing critical thinking, independent thinking and good judgement, rather than relying on AI for the answers?

Grace Chua: That concern is understandable. The key is to use AI in ways that extend a child’s thinking rather than end it. One of the most useful habits we can give children is to treat an AI response as a starting point for thinking, not the final word. That is a habit I hope to model for my own son too – being curious about an answer, but not simply handing over your judgement to it.

I like the simple family rule “attempt first, AI second.” Let the child form an idea, solve part of the problem or write a rough first draft, then use AI for a hint, another explanation, practice questions or feedback. That helps keep the effort and thinking with the child, while also making it easier for a parent to see where they may genuinely be struggling.

Parents can check part of the activity by asking, “How could we verify this?”, “What evidence supports it?”, or “Does another trusted source agree?” Asking our kids to explain an answer in their own words is especially useful. It shows whether they have actually understood it and returns the thinking to them. The goal is not to make children distrust everything AI says.

The same principle applies to creativity. Rather than asking AI to generate everything from scratch, children can use it to push an idea further like to explore another angle, identify something they may have missed, or test a counterargument. The child should still be the one deciding what to keep, reject, combine or rewrite.

The goal is not to make children distrust everything AI says. I would not want my son to grow up fearful of technology, just as I would not want him to trust it blindly. I would want him to feel confident using it while keeping his own judgement. AI can be a collaborator or a sounding board, but the effort and final responsibility / accountability should remain human.

4. Looking ahead, what knowledge, skills or mindsets do you believe today’s children should develop to thrive in an AI-enabled future, and how can parents play an active role in nurturing them?

Grace Chua:  Children will need AI literacy, of course, but I think the bigger picture is just as important. They will need curiosity, communication, creativity and sound judgement. They should be able to ask good questions, recognise uncertainty, protect their privacy and decide when a human relationship or expert is more appropriate.

One principle behind the new ChatGPT for Teen experience is that young people should learn not only how to use AI, but how to understand it, direct it, question it and create with it. That is the shift we should aim for, from passive users of technology to confident creators who can apply what they learn to real problems.

At the same time, distinctly human qualities such as empathy, resilience, collaboration and a strong personal voice will matter enormously. Those are the qualities that help children form their own perspective rather than simply accepting what technology gives them.

On the role of parents, I know many of us are on our own journey using AI at work and in our personal lives. We can be curious alongside our children, model how we check information, talk openly about mistakes and set boundaries that evolve as they grow. We do not need to become AI experts overnight. Preparing children for an AI-enabled future is a shared responsibility across parents, educators, communities and technology companies. Our common goal should be confident, independent thinkers who can use powerful tools with healthy boundaries.

5. On a more personal note, is there an experience or observation that has shaped your perspective on how families can embrace AI thoughtfully and responsibly?

Grace Chua: My son is 2.5 years old, so he’s certainly not old enough to be using AI himself yet! But the truth is, he’ll never experience AI as something new. By the time he’s old enough to use it, it may feel as ordinary to him as the phone he already recognises in my hand today.

What strikes me most as a parent is how early children start learning from our habits. My son already picks up a phone and pretends to talk into it, smiles when a camera is pointed at him, and turns around when he hears me speaking to ChatGPT in voice mode. Whenever he does that, it’s a good reminder to myself that kids do start learning from our habits long before they are old enough to use these tools themselves.

I’m conscious of this in my own life because technology is very visible to him. He sees me use it for work, to organise our lives, to learn things and sometimes simply to think something through. I want him to see that technology can be genuinely useful, but also that we remain the ones deciding when to use it, what to believe and when to put it down. I certainly won’t get that right all the time, but it has made me more aware that raising children to use AI thoughtfully starts, in some ways, with us.

I also think curiosity will matter even more in an AI-enabled world. When answers become easier to find, asking the right question becomes more important. I’d want my son to learn not just to ask AI for an answer, but to pause and think: What is the problem I’m actually trying to solve? What might I be missing? I want him to feel confident using technology, without assuming that a polished answer is always a complete one.

AI has been genuinely useful to me as a working mum. I’ve used it to bring a little more structure to juggling work, childcare and things I want to do for myself, or simply as a place to think something through. As my son gets older, I can imagine using it to help me explain something difficult or find calmer words when I’m tired and a conversation is hard. But I wouldn’t want to outsource the parts of parenting that really matter. AI can explain a lot of things, but it can’t make my son feel safe, loved or understood. It can’t replace the patience, judgement and presence that come from knowing your child.

And I think that’s the thing I keep coming back to. The tools our children grow up with will continue to evolve, but what they need from us probably won’t. They still need adults who are there, who listen, who guide them, and who help them develop their own judgement.

Interview with Sher-Li Torrey, Founder of Mums@Work

1. As both a working mother and the founder of Mums@Work, what are some of the most common questions or concerns you hear from parents about children using AI, and how can families approach these conversations with confidence rather than fear?

Sher-Li Torrey: I think it is important to recognise that these concerns are understandable, and a certain amount of caution is healthy. AI is developing very quickly, and we still do not fully understand how regular use may affect children’s learning, judgement and development. 

The areas of concern tend to fall into a few main areas:
(a) Will children stop thinking for themselves, become lazy or fail to develop their critical thinking skills?
(b) Does using AI count as cheating for school assignments?
(c) What age is appropriate to start, and will kids who do not use it fall behind their peers?
(d) What information are they sharing online?

At the same time, fear alone may lead to blanket bans without meaningful conversation. Parents have every right to set limits, delay access or decide that certain activities should be completed without AI. However, our children will grow up studying and working with these tools, so they also need to learn how to use them responsibly.

1. Families can start by having open conversations about how and why their children are using AI. Ask them to show you what they are doing, what prompts they used and how they decided whether the answer was accurate. Children should understand that AI can help them generate ideas, explain difficult concepts or improve their work, but it should not replace their own thinking.

2. Set clear family rules around privacy, fact checking, schoolwork and appropriate use. Discuss some of these points with them, especially the teenagers, as they navigate AI usage.

3. Know that you do not have to be an AI expert before starting these conversations. We can learn alongside our children, remain curious and help them develop the judgement to use AI well.

Confidence does not come from knowing everything about the technology. It comes from being clear about the values and abilities we want our children to develop, and then deciding whether AI supports or weakens them.

The aim is not to raise children who never use AI. It is to raise children who know when to use it, when to question it and when to think for themselves.

2. Many parents are still learning about AI themselves. What practical advice would you give parents who want to guide their children in using AI responsibly, even if they don’t consider themselves particularly tech-savvy?

Sher-Li Torrey: You don’t need to understand how AI works to parent well around it. Our role is not necessarily to know more than they do. It is to ask the right questions, set clear boundaries and help them develop good judgement.

Some practical ways to do this:
1. Use an AI tool together. Maybe ask our child to show you how they use it, what they use it for? E.g. when they create an image with AI – what was the purpose of the image? Why did they feel they wanted to use AI instead of drawing it himself/ herself? Or ask a question and discuss the response together. Ask ‘concept’ questions:
● Does this sound accurate?
● Can we verify it?
● Is this work/ answer entirely AI or have you (your child) done some of the thinking too?

2. As mentioned in Q1, family should agree on some basic rules for safety:
● Do not share personal information, photographs, passwords and personal family details
● Always verify important information using reliable resources
● Be honest when AI has helped with their schoolwork

I also encourage parents to focus on the purpose behind the use. Remind your children that AI can be helpful for explaining a difficult concept, generating ideas or giving feedback on your concepts. BUT if a child uses it to avoid thinking, reading or forming their OWN opinion, then it is not supporting their learning anymore.

Remember: Technology will continue to change, but the values we want to teach our children, remain the same: curiosity, honesty, responsibility and critical thinking. Parents already know how to guide their children in these areas, even if they are still learning about AI themselves.

3. Healthy digital habits begin at home. What simple routines or family practices would you recommend to help children balance AI use with independent thinking, creativity and meaningful offline experiences?

Sher-Li Torrey: In addition to what I have mentioned above, I would also add some points on developing better digital habits:

1. Have ‘protected offline time’ each day – eg. During meals or an hour before bed. Set these as ground rules for all family members
2. Adopt a “try it yourself first” rule: Always try to use your brain, rather than turn to AI first (EveniIf it feels easier)
3. Encourage physical, non-digital habits on a daily basis – sports, cooking, music. Something that does not need them to be online. That way their physical development is not compromised
4. Include your own usage of AI experience openly with them. (this I practice myself – because it encourages them to understand what limitations AI can or cannot have)

4. AI has enormous potential to support learning, but it also raises questions about academic integrity and over-reliance. How can parents encourage children to use AI ethically while still developing resilience, curiosity and problem-solving skills of their own?

Sher-Li Torrey: Parents could first help children understand the difference between using AI to support their learning VS using it to complete the work for them. If a child cannot explain the answer, defend the conclusion or show how they arrived at it, then they probably have not truly learned from the process. (Helping them recognise where the thought process happened for them is quite key in this case).

One practical approach is: Ask children to attempt the task first. (So, they can form their own ideas, make notes or produce a first draft before using AI to clarify a concept, challenge their thinking or receive feedback. Not every assignment or difficult question needs AI. Some could be a test of what they remember vs a deep thought-process required answer.

Resilience and problem solving develop when children have the opportunity to struggle, make mistakes and try again. We should not allow AI to remove every frustrating or difficult part of learning. Sometimes the most responsible and useful choice is not to use it at all.

In this aspect, parents can also ensure there is a culture of “failure or not knowing the right answer the 1st time” is NOT a bad thing. That kind of allows that resilience, curiosity and problem solving to have room to develop in our children. If every question had a perfect answer – then in some sense, the child’s brain does not get to understand why something is wrong, and how to ‘correct it’ – an essential life skill.

5. If you could share one key piece of advice with parents raising children in today’s AI-driven world, what would it be, and why do you believe it is especially important?

Sher-Li Torrey: My one key piece of advice would be to pay attention to how we ourselves are embracing AI in our
lives. What we ask of our children, we must also be willing to practise ourselves.

Integrity, critical thinking and responsible use are not standards that apply only to children. If we rely on AI for every answer, accept what it tells us without questioning it or use it to avoid thinking, our children will notice.

Our ability to think needs to be exercised. AI may help us work faster, but speed should not alwaysbe the goal. Both parents and children need to keep learning, questioning and adapting as the technology evolves.

Perhaps the most powerful thing we can do is model the relationship with AI that we hope our children will develop: curious but not dependent, open to learning but willing to question, and able to recognise when we should think for ourselves.

6. On a more personal note, you’ve spoken publicly about using AI with your own family. Could you share how AI fits into your home life, and are there any conversations, routines or experiences that have helped your children develop healthy and responsible AI habits?

Sher-Li Torrey: My elder child is a teenager, and we allow her to use AI. For one school presentation on sustainability, she first researched some ideas and then used AI to generate more sub ideas. She was quite pleased with the result and thought her idea was brilliant and very unique.

She was the third student to present. After listening to the first two presentations, she realised that both students had come up with almost the same idea. She was quite shaken because the idea she thought was so original was clearly not unique at all.

At that point, she realised that she had to use her own way of presenting the idea if she wanted to stand out. When she came home, we had a very interesting conversation about what the experience probably meant.

AI can help everyone produce better answers and more polished presentations. But if everyone is using similar tools, many people may also arrive at the same ideas. The people who stand out will be those who can think beyond the first answer, bring in their own perspective and express the idea more creatively.

With teenagers, I do not think it is realistic or helpful to supervise every prompt they type. Our routines are therefore based more on open conversation than policing. We encourage them to begin with their own ideas before using AI. We also ask questions such as: What did AI give you? What did you add or change? Can you explain the final answer in your own words? If factual information is involved, we remind them to check it against other sources.

For me, that experience taught her far more than simply telling her not to depend on AI. She discovered for herself that AI may help you reach a certain level, but it cannot replace your own thinking, creativity and voice.

The perspectives of Grace Chua and Sher-Li Torrey point to a balanced approach to parenting in an AI-enabled world: helping children use AI thoughtfully while developing their own judgement, creativity and critical thinking.

AI will continue to evolve, but children will still need guidance, curiosity, resilience and the confidence to think for themselves. Ultimately, helping children use AI well is about ensuring that technology supports their learning without replacing their own voice.