Developing AI literacyin primary learners.

Artificial intelligence (AI) is no longer a futuristic dream; it’s woven into the fabric of kids’ everyday lives. Think about it: they’re chatting with voice assistants, creating images with online tools, and getting video recommendations on YouTube, often without a second thought that they’re interacting with AI. As educators, this raises an urgent and timely question: when should we start teaching children about AI, and what should that look like in their early years?

Consider the findings from Ofcom’s Online Nation 2024 report, an annual study by the UK’s communications regulator that provides a comprehensive overview of online habits and trends. This report revealed that over half (54%) of British children aged 8 to 15 have used a generative AI tool in the past year. ChatGPT leads the list, with 37% of online children in this age group reporting its use, followed by Snapchat’s My AI at 30%. While many children turn to AI for fun, over half say they use it to support their studies. Perhaps most revealing is that one-third of these young users trust the information AI gives them. That statistic alone should make us pause and act.

This article explores why AI literacy is essential for primary learners and how we can start teaching it in developmentally appropriate, playful, and engaging ways. The aim isn’t to overwhelm young children with prompt creation techniques or increase screen time, but to spark curiosity, build awareness, and empower them to better understand the systems shaping their experiences. Because if children are already using AI, the real question is: how can we help them do it wisely?

Why bother with AI literacy in primary school?

As kids spend more time interacting with AI, often before they even understand what it is, the real question isn’t whether we should teach them about it, but how and when. Primary school might sound a bit early to dive into algorithms or machine learning. But it’s the perfect time to lay the groundwork for helping children understand, question, and use the technology that’s already shaping their world.

Let’s break down why it matters:

Making sense of technology

Ever heard a child wonder, “How does YouTube always know what I want to watch?” or “My music app just gets me — how does it pick songs I like?” These questions show that kids are already noticing patterns in how technology responds to them. Teaching AI literacy at this stage helps them realise that these tools aren’t magical — they’re built by people, trained on data, and shaped by human decisions. When children understand that tech is designed, not mysterious, they become more curious, confident users.

Learning to think critically and ethically

AI opens the door to important conversations about fairness, bias, privacy, and decision-making, all in a way that makes sense to young learners. When we help kids see that AI tools are collecting information to “make choices,” we can ask powerful questions together: “Where is this information coming from?”or “Is this always fair or right?” These are natural springboards into critical thinking and ethics, and children are far more ready for these conversations than we often give them credit for.

Using tech responsibly

We’ve all seen it: the quick click on “I agree” without even blinking. By building AI literacy, we help children slow down and think: “Should I be sharing this?” or “Who might be reading what I write?” Early discussions around digital safety, privacy, and boundaries set the stage for responsible, informed tech use down the road.

Avoiding misunderstandings and misuse

If kids don’t understand how AI works, they’re more likely to either misuse it or believe it without question. This isn’t about scaring them, it’s about giving them the tools to think critically, explore safely, and know when to ask, “Does this seem right?”or “Can I trust this?” With some basic AI literacy, children don’t just become safer users; they become empowered ones.

And here’s the thing: kids want to understand how AI works. Initiatives like the Children’s Manifesto for the Future of AI show that young people aren’t just curious, they’re asking real questions about trust, fairness, and control. The manifesto was developed from ideas, pitches and messages submitted to world leaders by children in the run-up to the first-ever Children’s AI Summit, which took place in the UK in February 2025, and it shows that these kids are not passive users. They’re ready to be active participants in shaping the future of technology.

FINDING THE SWEET SPOT: AI AND PLAY

Kids are curious about AI, but they also need time to play freely, invent games, tell stories, and explore without rules. That kind of unstructured play is essential for their development. So, how do we bring AI into the picture without taking that away?

The key isn’t more screen time or rigid lessons. It’s about blending AI into what kids already do. A quick game of “human robot” can spark conversations about how machines follow commands. A comment about YouTube suggestions can lead to a chat about algorithms. These moments don’t need tech, just curiosity and good timing.

We also need to meet kids where they are, using simple language and keeping things creative. And yes, many adults are still figuring out their stance on AI — that’s okay. With care, we can introduce AI in playful, natural ways that feel right for young learners.

KEY CONCEPTS FOR AI LITERACY

When we talk about AI literacy for primary learners, we are talking about demystifying the technology and fostering foundational understanding. Educators should focus on these key concepts to build AI literacy:

AI as a Human-Made Tool: Beyond Magic

AI is a sophisticated tool created by humans, not a mysterious entity. We need to empower children to see themselves as potential innovators. We can frame AI like other familiar tools (e.g., a tablet or a smart appliance), emphasizing its design and purpose.

Decision-Making in AI: The “Sense-Think-Act” Cycle

Many AI systems operate through a cyclical process of collecting data, processing it, and generating an output. This simplified model (Sense → Think → Act) provides a concrete framework for understanding AI functionality. It introduces basic computational thinking concepts like input, processing, and output in an accessible way.

• Sense: AI gathers information (data).
• Think: AI processes data using programmed instructions.
• Act: AI produces a result or takes an action.
• AI vs. Human Thinking: Distinct Forms of Intelligence

AI’s “intelligence” is computational and task-specific, fundamentally different from human consciousness, emotions, or intuition. This distinction reinforces that AI does not “think” or “live” like humans. It emphasizes the unique value of human qualities such as creativity, empathy, and critical judgment.

AI’s Problem-Solving Abilities: A Powerful Partnership

AI excels at assisting humans with complex problem-solving by processing large datasets and identifying patterns. AI can be very useful as an augmentative tool, not a replacement. We should demonstrate how AI can enhance human capabilities while reinforcing that the ultimate responsibility for ethical problem definition and solution lies with humans.

AI and Responsible Use: Guiding Young Learners

Engaging with AI tools involves considerations of safety, privacy, and age-appropriateness. This promotes early digital citizenship. Educators should emphasize the role of adults in ensuring children interact with safe, age-appropriate AI applications. It’s important for children to understand the importance of always asking a trusted adult before using new online tools. This fosters critical evaluation of online resources and safe digital habits.

IMPLICATIONS FOR TEACHERS

As generative AI tools become increasingly accessible, it’s important to note that most platforms enforce age restrictions (typically 13+), due to the risks of unfiltered content. For primary learners, thus, AI education shouldn’t begin with text or image generators—instead, we must prioritize hands-on, age-appropriate activities that demystify how AI works without exposing children to complex ethical dilemmas or unreliable outputs. Let’s move on to check some possibilities!

Quick, Draw!

Quick, Draw! provides a tangible example of AI in action. Children can see that the AI is not “magic” but a tool trained to recognize patterns (e.g., how a cat or a tree is typically drawn). Teachers can show students the dataset and explain: “The AI learned from thousands of drawings made by people like you. It uses that knowledge to guess what you’re drawing!”

Teachable Machine

This is a machine learning experiment, where children can actively participate in teaching the AI by providing examples (e.g., showing a pencil multiple times). This demystifies how AI “learns” from data and illustrates the “Sense-Think-Act” Cycle:

• Sense: The AI “sees” images or “hears” sounds (input).
• Think: It compares new inputs to the trained examples (processing).
• Act: It responds with a label or action (output), like saying, “This is a pencil!”

Say What You See

In this experiment, kids are invited to play with the art of image prompting. They see an image and have to create a prompt to try to generate an image as close to the original as possible. This activity helps students understand image generation and see that there’s no magic to AI: with the right prompt you can generate/recreate images. This experiment can be a great collaboration task – when students collaborate to generate the best prompt possible, team work pays off! Each kid will notice a different detail!

Odd one out

Is it possible to distinguish real artwork from AI-generated artwork? This is what this experiment proposes. Kids are invited to guess the AI-generated one, and as they move on, they can share the details that helped them guess correctly. This activity can also spark a discussion on AI and copyrights: Who owns an AI-generated image or artwork?

Day of AI

Day of AI offers a comprehensive curriculum for AI literacy developed by leading faculty and educators from MIT RAISE. The “What is Artificial Intelligence?” lesson for primary learners helps demystify AI with Familiar Analogies by comparing AI to tools like magnifying glasses or calculators. Students are also invited to act as “AI trainers” and experience the “sense-think-act” cycle.

Unplugged AI

Many AI tools require digital interaction, but young children benefit from tactile, social learning. Did you know that there are many unplugged activities that can teach AI? Yes, you do not necessarily need a computer, experiment or tool in order to teach AI to your students. This booklet of Unplugged Activities provides approaches that help learners of all ages to experience AI actively and without the use of a computer.

Conclusion: Preparing Kids for an AI World

AI isn’t coming—it’s already here, shaping how children play, learn, and explore. With kids interacting more and more with AI, we can’t wait to teach them how to engage with technology thoughtfully.

The answer isn’t more screen time or complex lessons, but playful, age-appropriate experiences that build awareness, not dependency. Through activities like training “paper AIs,” spotting AI-generated art, or discussing why tools make mistakes, we help children:

• See AI as a human-made tool, not magic
• Ask critical questions about how it works
and who controls it
• Develop digital discernment alongside curiosity

This foundation prepares them to use—and question—AI wisely, both today and in the future. The time to start is now. Key Takeaways:

• Start early with play-based AI literacy
• Focus on understanding over usage
• Empower kids to be critical thinkers,
not passive users

By demystifying AI today, we give children the tools to shape tomorrow’s technology responsibly and confidently.

THE AUTHORS:

Roberta Freitas is an Educational Technology specialist with over 15 years of experience in language teaching and over 10 years as an Educational Technology Coordinator. Holding an MA in Language Studies (PUC-Rio), she is a Google Certified Trainer, Google Innovator, and lead of the Google Educators Group Rio de Janeiro. Currently, she focuses on AI in education, delivering teacher training and research that bridges language pedagogy, EdTech, and ethical AI integration.

Isabella Campos is a graduate teacher who specializes in language education. She has extensive experience as a teacher, trainer, and academic coordinator and works with a strong focus on young learner education, bilingual programmes, and academic management. She holds a Bachelor’s degree in Languages, a Diploma in Bilingual Education, and a Master’s in Professional Development in Language Education from the University of Chichester
in the UK. She is currently pursuing a postgraduate course in Early Childhood Education at PUC-Rio.