Make
Seven AI activities for middle and high school students
Try seven practical activities that help students explore how AI works, question its outputs and think about its effects.
Boundlearn Editorial
Learning about artificial intelligence can start with paper, a pencil and a question. Students can explore how people organise examples, test decisions and check claims before they use a chatbot or write code.
The activities below are starting points for families and educators. The suggested ages and times are estimates, not fixed requirements. For online tools, an adult should check the service's current age rules and privacy settings first. Use made-up examples and avoid sharing personal information.
Activities for ages 10–14
1. Sort a set of imaginary objects
Time: About 15 minutes. You need: Paper and a pen.
Draw twelve imaginary creatures with different shapes, colours and numbers of legs. Give each a name, then ask someone else to sort them into two groups and explain their rule. Add a creature that doesn't fit neatly and see what happens.
This is a human sorting game, not an AI model. It introduces the idea that categories depend on the features and examples people choose. Ask: could a different rule produce a different result?
2. Design a decision tree for a rainy day
Time: About 20 minutes. You need: Paper, a pen and a few imaginary scenarios.
Write a question at the top of a page: “Is it raining?” Draw branches for yes and no. Add a second question to each branch, such as whether an activity is indoors, until you reach a suggestion. Then try a scenario the tree didn't anticipate.
The exercise gives students a simple way to describe decisions as a sequence of rules. It doesn't recreate every kind of machine-learning system. Discuss where the rules came from and when a person might want to override them.
3. Spot the gaps in a set of examples
Time: About 20 minutes. You need: A set of drawings or descriptions.
Imagine you are making a picture-sorting system that needs to recognise bicycles. List ten examples you might include. Are they all adult road bikes seen from the side? Add examples with different sizes, settings and viewpoints, then list situations the set still misses.
This is a planning exercise; you don't need to train a model or upload pictures. It opens a conversation about how the examples used in a system can affect what it handles well. Experience AI's foundations unit includes a lesson on how a limited dataset can lead to bias.
4. Map an AI decision and who it affects
Time: About 15–25 minutes. You need: Paper and a pen.
Imagine a library app that recommends a book. Draw three boxes: information the app might use, a recommendation it might make, and a person who receives it. Add questions around the diagram: What if the information is incomplete? Can the reader choose something else? Who decides what the app can see?
There is no single correct diagram. The point is to make the assumptions visible enough to discuss. For a guided unit, Experience AI offers lesson plans for ages 11–14.
Activities for older students
5. Check a confident-sounding claim
Time: About 20–30 minutes. You need: A claim and access to reliable sources.
Choose a short factual claim from an article or an AI-generated answer. Identify the original source, locate the relevant passage and write down whether it supports the claim. Note any missing context or uncertainty.
If you use a generative AI tool to produce an example, first check its age and account requirements and use it under appropriate adult supervision. Do not enter private schoolwork, personal details or information about other people. A convincing answer is a reason to check the evidence, not a substitute for it.
6. Compare how two resources explain AI
Time: About 30–45 minutes. You need: Two educational resources and a comparison table.
Choose a shared topic, such as training data or how an AI system can make mistakes. Compare which examples each resource uses, who its intended audience is and what questions it leaves unanswered. Finish with a short explanation in your own words.
Two places to begin are Stanford CRAFT, which offers adaptable AI-literacy resources for high-school educators, and Day of AI, which offers classroom resources for different age groups. Check the individual activity's requirements before using it.
7. Write a brief for a useful AI-assisted tool
Time: About 30–45 minutes. You need: Paper or a document.
Pick a small problem, such as helping a library organise a list of fictional book descriptions. Describe the intended user, what information the tool would receive and what a useful result would look like. Add three examples where its output should be checked by a person.
Finish by asking what would happen if information were missing, inaccurate or private. You can complete this as a paper design; building a working AI system is a separate, larger project.
Where to go next
Try one activity first, then use the questions it raises to choose the next. A student who enjoys testing examples might explore data and coding; someone who likes discussing consequences might look at ethics, design or policy.
SparkLab Junior is Boundlearn's online academy for ages 10–14, using AI as a lens for problem-solving, critical thinking and responsible decision-making. For older students interested in building technical projects, explore the specific Delta AI pathways and check their current eligibility and curriculum.
Sources and further reading
Reviewed 9 November 2026. Official guidance can change; follow the source links below.