EU AI Act risk classifier
Twelve questions, in the order the Regulation asks them. Every verdict shows the articles behind it, so you can check the reasoning rather than trust the output. Runs entirely in your browser, nothing is transmitted.
How this works
The classifier applies the tests in the order the Regulation applies them. Getting the order wrong is the most common source of wrong answers: a prohibited practice is prohibited whether or not it also appears in Annex III, and Article 50 attaches regardless of risk tier.
- Article 3(1): is it an AI system? The test is inference. Definition →
- Article 2: is there an EU nexus, including output used in the EU? Scope →
- Article 5: is the practice prohibited? Article 5 →
- Article 6(1) / Annex I, then Article 6(2) / Annex III: two routes, two dates. High-risk →
- Article 6(3): is the derogation available, and does profiling remove it? Derogation →
- Article 50: transparency, at any tier, live now. Article 50 →
What this is not
This is not legal advice and it is not a conformity assessment. It applies the Regulation's tests to the answers you give. Classification depends on the intended purpose of the specific system and on facts a questionnaire cannot capture. Article 6(4) requires a provider who concludes an Annex III system is not high-risk to document that assessment before placing the system on the market, and Article 80 lets an authority challenge it. Prepare that assessment with qualified counsel. Our editorial standards →