Harvey's AI Compliance Tool
Find out in 5 minutes which AI Act risk category your system falls into
A free, enterprise-grade self-assessment tool that examines your corporate AI system across 6 sections, 19 questions and 5 dimensions — with a live dashboard and a downloadable, detailed PDF result.
How it works
Three steps to your classification
No account, no data transfer. The answers describing your system stay in your browser.
Identify the system
Enter the AI system name, a short description and the industry context. These also appear on the PDF result for sharing within your team.
19 questions, live dashboard
We guide you through 6 sections. Your score and category estimate update continuously, and the edge case detector flags contradictions.
Detailed result + PDF
You get the risk category, the specific obligations, a scenario simulator, and you can download the branded PDF summary.
The downloadable result
A PDF you can forward to decision-makers
- A short, clear explanation of the risk category
- A concrete obligations list specific to your system
- Per-dimension scoring and all of your answers
- A "what if…?" scenario simulator to reduce risk
- Harvey's branding and a direct, clickable contact
The four risk categories
Where could your system fall?
Minimal risk
Most business AI. No specific AI Act obligations, but AI literacy training is mandatory for everyone.
Limited risk
Chatbots, generative content. Transparency obligation: the user knows they are dealing with AI.
High risk (Annex III)
HR, credit, education, biometrics. Full compliance by 2 Aug 2026 — documentation, oversight, audit.
Prohibited practice
Social scoring, manipulation. Prohibited under Article 5 — the use must be discontinued.
Why now
The AI Act is fully applicable from 2 August 2026
High-risk system obligations take effect. The first step is always to clarify whether — and to what extent — you are affected.
Ready? Find out in minutes.
Start the risk classification — free, private, and you get a tangible result at the end.
Start assessment