Methodology · Game-based learning
Serious games for AI regulation
Serious games are games designed for learning rather than pure entertainment. This page explains why the format suits regulation — a domain usually taught as text — and where its limits are.
01Why simulation fits complex rules
Regulation is conditional logic applied to messy facts. Texts teach the logic; only situations teach the application. A simulation supplies endless situations at zero stakes: learners apply the rule, fail safely, and retry — the same reason flight simulators and moot courts exist.
02Scenario-based reasoning
In a scenario, concepts stop being definitions and become tools. "Deployer responsibility" is inert in a slide; it becomes vivid when you must decide who answers for a biased hiring filter: the vendor who built it or the office that used it unread. Scenarios also carry ambiguity honestly — most real classification questions are arguable, and a good scenario preserves that.
03Decision-making under uncertainty
Real inspectors never have complete information. Deciding with partial evidence — and being accountable for the decision — is itself the competence. A game can force that experience: in NO AI ACT you must classify with the dossier you have, cite your evidence, and watch the decision hold or collapse.
04Honest limitations
- A game teaches logic and habits, not law: no simulation substitutes for the text or professional judgment.
- Simplification is the price of playability: every case compresses reality.
- Motivation varies; games are a format, not magic.
- Learning-impact claims require studies. This project makes none: its effectiveness has not been empirically validated.
05How NO AI ACT applies this
Thirteen cases spanning the risk ladder, evidence-grounded decisions, a reflective debrief after every report and visible consequences on a simulated city. Design details are on the game's rationale page; classroom scaffolding is in the teacher guide and activities.