What the tools do, and what they do not
The tools give a first direction in a few minutes: which EU AI Act risk tier your system falls into, which obligation that triggers on which date, and the penalty ceiling. Your answers are calculated in your browser and are not sent to a server.
The result is not legal advice. A definitive classification requires a written review of the system's purpose, where it is used and your role (provider or deployer).
After you get a result
If the system comes out high-risk or within the transparency duties, the next step is an inventory: every AI system in the organisation, including purchased tools. Detailed dates are in the compliance timeline guide; the build-out work is described under AI compliance and governance.
Risk tiers at a glance
| Tier | What it means | Date |
|---|---|---|
| Prohibited | Practices listed in Article 5; may not be used | Since 2 February 2025 |
| High-risk | Risk management, data governance, technical documentation, human oversight | Annex III: 2 December 2027 · Annex I: 2 August 2028 |
| Transparency (Article 50) | Disclosure of AI interaction and of AI-generated content | 2 August 2026 |
| Minimal risk | No tier-specific obligations; Article 4 literacy measures apply | — |