arqmetrica
Free tool

EU AI Act Risk Classifier — which tier is your AI system?

The EU AI Act (Regulation (EU) 2024/1689) sorts every AI system into one of four risk tiers, and the tier decides what you must do before 2 August 2026. This free classifier walks you through the plain-language triggers for each tier and returns the concrete obligations that attach. Tick what your system does, hit classify, and read off where you stand — prohibited, high-risk, limited (transparency), or minimal.

This classifier is guidance, not legal advice. It simplifies Regulation (EU) 2024/1689 to help you orient quickly — the legal text, delegated acts and national guidance govern. For anything that could be high-risk or prohibited, get a qualified assessment before you ship.

Describe your system

Tick every statement that is true of the AI system you are assessing. Leave the rest blank. The classifier reads the tiers in priority order — a single prohibited or high-risk trigger decides the outcome.

Prohibited practices (Art. 5)

Banned across the EU since 2 February 2025. If any of these apply, the system cannot be placed on the market or used at all.

High-risk uses (Annex III / Art. 6)

Permitted, but subject to the full compliance regime from 2 August 2026.

Transparency triggers (Art. 50)

Not high-risk on their own, but they create a duty to tell people that AI is involved.

Questions people ask
When do the EU AI Act obligations start?
Prohibited practices (Article 5) and the AI-literacy duty (Article 4) have applied since 2 February 2025. Rules for general-purpose AI models and the governance framework started on 2 August 2025. The full high-risk regime for Annex III systems and the Article 50 transparency rules apply from 2 August 2026. High-risk systems that are regulated products under Annex I follow on 2 August 2027.
What is the difference between high-risk and limited-risk AI?
High-risk systems (Annex III / Article 6) are used in sensitive domains — recruitment, credit, education, biometrics, essential services — and must meet the full compliance regime: risk management, data governance, technical documentation, human oversight and conformity assessment. Limited-risk systems only trigger transparency duties under Article 50: telling people they are dealing with AI and labelling AI-generated content.
Is this classifier legal advice?
No. It is a fast orientation tool based on Regulation (EU) 2024/1689. It does not account for delegated acts, national implementation or the specifics of your deployment. Treat a high-risk or prohibited result as a prompt to get a qualified legal and technical assessment.
Next step

Not sure the mapping holds for your system?

An AI Act audit pins down the tier, the gaps and the evidence you need before 2 August 2026 — with a documented trail your enterprise customers and auditors will accept.