ImplementationExplainer

AI Literacy for Students: What to Teach Before Independent Use

What do students need to learn before using AI on their own? Start with seven skills, from checking an answer to knowing when to leave AI out.

Illustrative scene of a teacher and two students examining printed material together
On this page5 sections
  1. Learning comes first
  2. Check AI answers
  3. Privacy and honesty
  4. Work without AI
  5. Start with one lesson

AI literacy for students is knowing how AI can help with schoolwork, where it can go wrong and what to check before using an answer. Students also need to know what information to keep private, how to explain the help they received and when to do the work without AI.

These lessons can begin with a teacher’s demonstration or a printed example. Students don’t need their own AI accounts to take part.

For schools planning what to teach, seven skills provide a useful starting point:

  • Set a purpose: know what you are trying to learn.
  • Understand the tool: know why a convincing answer can still be wrong.
  • Check the answer: look for evidence before relying on it.
  • Notice bias: ask whose views or experiences are missing.
  • Protect information: follow the school’s rules about what can be shared.
  • Explain AI’s help: say what it contributed and what you did yourself.
  • Choose when to leave AI out: recognise tasks that need your own thinking.

This is TopSchool’s selection for teaching before independent use, drawing on the OECD and European Commission’s AI literacy framework. A school will still need a wider curriculum and its own rules about which tools students may use.

Start with what students are learning

Take a lesson on evaporation. If students need to explain the process, that explanation should still be theirs. A teacher-provided text might help them work out unfamiliar words, while they do the thinking needed to connect the ideas.

Before introducing AI to a task, make that distinction clear. What can it help with? What must students work through themselves?

Students also need a simple explanation of how the tool works. A system that generates text uses patterns learned from data to produce an answer. It can put words together convincingly and still get the facts wrong.

You could show the class a well-written AI answer containing an error. Ask students to find the mistake and explain why the rest of the answer made it easy to miss. That gives the class something specific to discuss before anyone relies on a similar response.

UNESCO’s student competency framework brings together understanding AI and keeping people in control of its use. In a lesson, those ideas belong together: students need to understand the tool and the learning they remain responsible for.

Show students how to check an AI answer

Give students a short statement about a local environmental issue and a relevant source. Ask them to find the part of the source that supports the statement. Does it say what the AI answer claims, or has something been added or left out?

They might decide to keep the statement, correct it or set it aside until they know more. The AILit Framework’s guidance on evaluating AI answers connects checking an answer with deciding whether to use it. Asking another chatbot the same question doesn’t replace looking at the evidence.

Then look at the viewpoints in the answer. Two accounts of the same issue may include accurate facts but leave readers with different impressions. Compare whose concerns each account includes and whose it leaves out. This gives students a way to discuss bias without expecting them to judge an entire AI system.

The useful part of the exercise is the student’s explanation: what they checked, what they changed and what they are still unsure about.

Teach what to keep private and what to acknowledge

Students need clear rules about sharing information, with someone they can ask when they’re unsure. UNICEF’s guidance on AI and children makes privacy, safety, fairness and accountability part of the wider responsibility for children’s use of AI. The school remains responsible for those protections.

Practise with fictional material. Use a made-up class project, a personal message and a record about another person. Ask students which examples the school’s rules allow them to share, which to keep private and which they would ask an adult about. Explain that removing a name may still leave details that identify someone.

Tell students who can answer those questions and whom to contact if information has already been shared. The school’s AI responsibilities and decision-making roles should make it easy to get help.

Students should also be able to explain AI’s part in their work. In a writing task where AI is allowed, ask what it helped with, what the student kept or changed and what they checked. UNESCO’s guidance on marking AI-generated content supports teaching students to make that help clear.

Keep the explanation suited to the task and the school’s assessment rules. Knowing how AI helped is useful, but teachers still need to check the student’s understanding.

Make room for work without AI

Sometimes the point of a task is to find out what a student can explain on their own. In another lesson, the teacher may want the class to critique an AI-generated explanation. Students need to understand why AI has a different place in each.

Compare those two tasks with the class. Ask students to explain what they would lose if AI did the thinking for them, and what they might learn by examining an answer together.

Choosing to work without AI should be an ordinary option. If a use is unclear or outside the school’s rules, students should stop and ask. Being confident with a tool does not give them permission to use it for every task.

Begin with one lesson

Choose a lesson already in your plans and one skill to practise. For checking answers, for example, ask students to correct one claim and point to the evidence they used.

Adjust the support to the learners. Some may need help finding a claim to check. Others may be ready to compare it with a source you provide, or explain their decision with less help. Age and development matter, but the support needed can also change from one task to the next.

Use demonstrations or printed material so the lesson doesn’t depend on students having an AI account or access at home. Deciding whether they can use a tool independently remains a separate school decision.

After the lesson, review a few student explanations with a colleague. Where did students need help? Use what you find to choose the next thing to practise.

Planning how to teach AI literacy?

Bring one lesson and talk through what your students need to learn before using AI on their own.

Discuss your school's approach
Portrait of Jayden Duong, Community Leader at TopSchool
About the author

Jayden Duong

Community Leader, TopSchool

Jayden builds connections around the questions that matter to educators and school leaders exploring AI. As Community Leader at TopSchool, he turns shared experience into useful conversations, practical knowledge and lasting connections across the education community.

Community engagementDigital communicationContent discovery
View profile and articles