Kettera AI Architecture Practice

Classifier · runs in your browser

AI Use-Case Risk Classifier

Describe a use case below. You get back its EU AI Act tier, whether it crosses the DTA impact-assessment threshold, the Privacy Act ADM position, the obligations that attach, and a transparency-statement starter. The logic is deterministic and published; nothing you enter leaves this page.

No data leaves your browser Indicative — not legal advice
1 · Who is deploying it?
2 · What does it do?
3 · Does it involve any of these? (Article 5 — prohibited)
4 · Does it do any of these? (Article 50 — transparency)
5 · Optional — to seed the transparency starter

How it works

Deterministic classification. The logic is published.

The verdict is produced by a transparent rule engine, not a language model — a CISO should be able to see exactly why a use case was classified the way it was. The EU tiers are a conservative reading of the AI Act’s Article 5 (prohibited), Annex III (high-risk) and Article 50 (transparency) tests; the AU obligations mirror the regulatory canon (source-verified). Where this tool simplifies, it errs toward caution and tells you so.

The AI-drafted plain-language explanation and a polished transparency statement are a Phase-1 server addition; today the starter is generated from a fixed template so the tool stays fully client-side and private.