AI school project: connecting learning, data and digital creation

While a one-off AI activity tackles a specific task, a school project brings together several subjects and produces a tangible outcome that someone will actually use. This article explains how to design such a project — from choosing a topic and researching needs to prototyping and testing.

The difference between an activity and a school project

• Activity: one lesson, one task, one output for the teacher.
• Project: weeks or months, several stages, an output for a real target audience.
• Activity: pupils complete an assigned task. Project: pupils help shape the task.
• Activity: AI is a helper. Project: AI is one of the tools, not the goal.

How to choose a topic

A strong project addresses a problem that people actually experience. Look for topics within your school community: a school website, an AI guide to the building, a mobile app for parents, an educational game, a sustainability campaign. Avoid vague briefs (‘anything involving AI’) — without context, a project falls apart.

Cross-curricular connections

• Czech language: writing the brief, creating texts, critically reading AI outputs.
• Mathematics and statistics: collecting and interpreting survey data, creating charts.
• Computing: prompts, working with data, creating a prototype, verifying sources.
• Design and art: visual style, illustrations, presentations.
• English: researching international sources, communicating with AI in English.
• Citizenship: ethics, personal data protection, copyright.

Researching needs and working with data

Before the team starts creating anything, it needs data. A short survey of the target audience (10–20 respondents) is usually enough. AI can help formulate questions, organise responses and identify patterns — but pupils make the final interpretation, because AI does not know the school's context.

Copyright, privacy and responsibility

What a school prototype should include

A prototype does not have to be a finished product — but it must show how the solution works. A clickable wireframe is enough for a website, a set of screens for an app, and a description of the rules plus one round for a game. Testing with three to five users outside the team will reveal most weaknesses.

How to build on the project over time

If the project captures pupils' interest, there are several ways to take it further: incorporate feedback, team up with another class, or present the outcome to school leaders. We describe a practical one-day version in the article AI project day at school.