Lesson plan: an introduction to the AI Act

01Audience, duration, objectives

• Target audience — upper-secondary (15+) and introductory university; no legal prerequisites
• Duration — 60 minutes core; 90 with the extension
• Objectives — students can state what the AI Act is, explain the risk-based approach in their own words, and justify a risk classification with facts from a scenario
• Preparation — play the first game case yourself (15'); skim the plain-language guide; have devices or the paper sorting cards ready

02Lesson flow

1. Hook (10') — describe the scored city (no regulation named): housing priority, queue order and benefits follow a single "reliability" score. Vote: acceptable or not? Collect reasons on the board.
2. Name the frame (10') — introduce the AI Act in three sentences: European regulation; risk-based; from banned practices to everyday uses. Show the four levels from risk categories.
3. Play (20') — in pairs, students play the game's first case (social scoring). Instruction: note which piece of evidence changed your mind.
4. Compare reports (15') — two pairs with different outcomes read their inspection reports aloud. Same evidence, different decisions: why? Anchor the debrief in the evidence, not opinion.
5. Close (5') — one look at the real text on EUR-Lex: the simulation ends where the law begins.

03Assessment and reflection

Exit ticket, three lines: (1) one practice the AI Act bans and why; (2) the evidence that mattered most in your case; (3) one question you still have. The third line seeds the next lesson.

04Extension activity (30')

Run the boundary debate from the activities page (credit score vs civic score) — it directly tests whether the risk logic survived contact with a grey case. Follow-up lessons: the risk-based approach and transparency and users.