V77 Trust
Start free trial

Classify a system's EU AI Act risk tier

How the classification questionnaire works, what the result means, and what to do if you disagree.

1 min read·Updated 8 November 2026Compliance leadsAI owners

How it works

Open a system and choose Classify risk. The questions follow the law's own order:

  1. 1
    Prohibited practices (Article 5): any "yes" makes the system prohibited.
  2. 2
    High-risk uses (Annex III) and regulated products (Annex I).
  3. 3
    The Article 6(3) exceptions, if an Annex III use applies but the system doesn't materially influence decisions.
  4. 4
    Transparency (Article 50) and general-purpose AI questions.

The result shows the tier and the reasons, quoting the questions that led to it.

What happens next

  • An assessment is created with the controls for that tier and your role.
  • The obligations page shows the dates that apply to the system.
  • If the system becomes high-risk, reviewers get a notification.

If you disagree

Re-run the classification with corrected answers; the previous result stays in the activity log. If you rely on an Article 6(3) exception, write your reasoning in the notes: you'll need it for registration and for any regulator who asks.

Watch out

The classification applies the legal tests to your answers. It is not legal advice, and the responsibility for the classification stays with you. For borderline cases, ask a lawyer and record their view as evidence.

Ready to try it in V77 Trust?Classify a system →

Related