Turn an AI use case into its full regulatory footprint — every domain it touches, from AI law and data protection to cyber, product safety and sector rules — with the obligations, the architecture and the evidence you owe, in about two minutes.
Community-curated knowledge graph — every claim carries its citation across law, engineering and governance. Every change traceable →
What should the AI system do, who does it affect, where will it run? Two or three sentences in English or German, or a project document. The graph then asks for the facts that decide the classification and reasons deterministically from there.
The analysis runs entirely in your browser — no text or document leaves this page.
Target marketsEuropean Union, United States (federal)Change
Next: the questions that decide the classification, then the result.
Or start from an example
Or upload a documentPDF, DOCX, TXT or MD — project sketch, tender document, architecture memo. Read in your browser only.
How this works, and what is logged
The knowledge graph reasons deterministically from your text, traversing the regulatory stack: detected attributes → regulatory footprint across all domains → AI-law dimension → articles → control objectives → evidence → vendor layers — and hands off to the business lens. Where a fact decides the classification and the wording leaves it open, the graph asks before it concludes.
What is loggedWhen you press Analyze, we log the description, your answers, the selected markets and the derived outcome to improve the graph. The log is anonymous: no account reference, no IP address and no session identifier is stored, and it stays anonymous even when you are signed in — a member reference is attached only if you tick the catalogue-contribution opt-in in the guided workflow. See the privacy notice.
Compliance by design = testing your AI system against the law before you build it
1Describe or test your ideaFree text or a project document — nothing is stored.
2See the full footprint + costsEvery regulatory domain triggered, obligations, evidence, cost bands.
3Adjust scope, oversight, designNarrow the purpose, add a human in the loop, change the architecture.
4Re-run in seconds — watch the verdict changeDeterministic: the same change always moves the answer the same way.
Design iteration costs seconds, not another legal review cycle — the loop runs against graph v2.21.0, in your browser.
The four steps form a loop: after re-running the analysis you return to step one and adjust the idea again.