Student AI Literacy Activities: 10 Ideas That Meet the Programme's Requirements

Last updated: 2026-07-22

'At least two activities to raise students' AI literacy and skills' is the most flexible of the five deliverables, and the easiest to turn into a one-off event done for the sake of it. Here are ten ideas with real learning in them, along with a way to turn an activity into report evidence.

What the programme asks of student AI literacy activities

Circular Memorandum 221/2025 asks participating schools to run at least two student activities that raise AI literacy and skills during the programme. Note the word literacy. An activity should not stop at letting students play with AI. Students should understand how AI works, know how to check its output, and think about the limits of using it. Being able to use it, judge it, and take responsibility for it: those three levels make a good yardstick when you choose an activity.

Ten student AI literacy activity ideas

1. AI fact-checking workshop: students ask AI about topics they know well, such as school history or local landmarks, then find the errors in the answers and trace the correct facts. They see for themselves how AI can be confidently wrong.
2. Prompt engineering contest: on the same task, such as writing an introduction for the school magazine, each group improves its prompt step by step, compares the differences in output, and works out what makes a prompt effective.
3. AI spotting challenge: mix real photos with AI-generated images, and human writing with AI writing, and have students tell them apart and note the giveaways, practising media literacy along the way.
4. Gamified AI learning course: an interactive game that teaches the principles and limits of large language models step by step, suitable for a whole class within lesson time.
5. AI ethics court: hold a mock hearing on a fictional case, such as AI-ghostwritten homework or a deepfake video, with students taking the two sides and arguing over who bears responsibility.
6. Cross-subject AI creation exhibition: students complete a piece of work with AI's help, such as an illustrated story, music, or a presentation, and must also show the creative process and explain who did what between human and machine. The explanation is the point.
7. AI class for older adults: senior students design the materials and teach older people to use AI tools to guard against scams. Teaching others is the deepest form of learning, and it counts as service learning too.
8. AI careers day: invite speakers from different industries to share how AI is changing their work, and have students interview them and produce a careers profile.
9. Campus AI charter: the student council leads the drafting of a school-wide code for student AI use, consults classes, and then issues it, bringing the ethics discussion down to actual rules.
10. AI science inquiry: use AI to help design and analyse a small investigation, such as measuring the campus microclimate, and have students compare the gap between AI's suggestions and the measured results.

Three principles for choosing an activity

First, include a checking step. An activity should have a step for testing AI output, and that is what separates literacy from mere use.

Second, produce something. A piece of work, a report, a charter, teaching materials all count. With something tangible, the report and the sharing sessions have something to show.

Third, finish it within lesson time where possible, or set a clear slot outside lessons. An activity that eats up too much time outside class often ends with only a few top students taking part, which drifts from the aim of reaching everyone. Of the two activities, at least one should run at whole-form or whole-class level.

How to record the results of an activity

The report needs more than photos of the activity. Set up three kinds of record in advance for each activity: participation data (numbers, classes, hours); samples of student work, keeping five to ten per activity; and a short before-and-after quiz or reflection survey of three to five questions to show whether understanding changed, for example whether fewer students agree that 'AI's answers are always correct'. With all three in place, one activity is one complete piece of evidence.

Common questions

Can the two activities be the same activity held twice?

The circular asks for 'at least two' student activities. Repeating the same activity in a different year or year level is, strictly speaking, still the same activity design. The safe approach is to prepare two activities of different kinds, for example a literacy course for a whole form within lesson time, plus an open-enrolment creation task or contest.

Which activities suit primary students?

Primary schools do better with activities that are concrete, visual, and have a clear right and wrong: the AI spotting challenge, the fact-checking workshop (using topics students know), and the gamified course all suit upper primary. Abstract ethics debates and charter drafting fit upper primary through secondary better.

Can an activity be outsourced to a supplier?

An external provider can lead it, and the funding can pay the cost. But the school should make sure the activity includes checking and reflection, and keep the student output as school-based evidence. A purely for-show AI experience day carries limited weight in the report.

Sources

EDB Circular Memorandum 221/2025 (Chinese)

Want to know your school's next step?

Tell us where your school is with the funding programme. We help with evidence, from choosing tools to filing the reports.

Book a free school consultation Request a proposal