EU AI Act risk classifier

The EU AI Act sorts every system into one of four tiers, and each tier carries different duties and a different deadline. Answer five questions about your system and get a plain-language read on where you stand: your risk tier, the specific articles that apply, a dated obligations checklist, and the date you need to be ready by. Everything runs in your browser. Nothing about your system is sent anywhere.

Step 1 of 5
What is your role for this system?

It changes which obligations land on you.

Get an email if the EU AI Act timeline changes.

One field: your email. Then confirm one link.

Free. We email you only when that page actually changes, at most one email per change. One-click unsubscribe, and we never share your address.

The four risk tiers

The Act (Regulation (EU) 2024/1689) is risk-based. The more a system can affect people, the more it has to do before it goes live.

  • Prohibited (Article 5). A short list of uses that are banned outright, such as social scoring, untargeted facial-image scraping, and emotion recognition at work. No safeguards make them lawful.
  • High risk (Article 6 and Annex III). Systems used in areas like biometrics, employment, education, credit and insurance, or critical infrastructure. They carry the full regime: risk management, data governance, technical documentation, human oversight, conformity assessment, and post-market monitoring.
  • Limited risk (Article 50). Systems that interact with people or generate content. The main duty is transparency: tell people they are dealing with AI, and label synthetic media.
  • Minimal risk. Everything else. No system-specific obligations, though the AI literacy duty (Article 4) and any voluntary codes still apply.

General-purpose AI is a separate track

If you train or substantially modify a general-purpose (foundation) model, you carry the GPAI duties in Articles 53 and 55 on top of whatever tier your applications fall into. That is why the questionnaire asks about it directly.

The deadlines that matter

  • 2 February 2025: the prohibited-practice ban and the AI literacy duty already apply.
  • 2 August 2025: GPAI model obligations, governance, and penalties apply.
  • 2 August 2026: the Article 50 transparency duties apply. This date did not move.
  • 2 December 2027: the Annex III high-risk duties apply. A political agreement of 7 May 2026, the digital omnibus, pushed this back from the original 2 August 2026.
  • 2 August 2028: the extended deadline for high-risk systems that are safety components of products under existing EU product law (Annex I), also moved by the omnibus from 2 August 2027.

The original dates come from the Regulation itself; the high-risk deadlines above reflect the May 2026 omnibus, which still needs formal adoption. They can move again through Commission guidance or amendments, so the tool links to the official page and can email you if that page changes. For the detail, see our note on the postponed high-risk deadline.

This is a starting point, not legal advice

The classifier mirrors the structure of the Act so you can see where you likely sit and what work it implies. Edge cases, exemptions, and the precise scope of Annex III are matters for qualified counsel. Use the result to scope the conversation, not to replace it.

EU AI Act questions, answered

What are the EU AI Act risk tiers?

Four: prohibited, high-risk, limited-risk (transparency) and minimal. The classifier walks five questions to place your system, then lists the exact articles that apply and the deadline that matters.

When does the EU AI Act apply to me?

It depends on your tier: prohibited practices since February 2025, GPAI duties from August 2025, Article 50 transparency duties from 2 August 2026, and most high-risk obligations from 2 December 2027 after the May 2026 digital omnibus pushed them back (product-embedded high-risk from 2 August 2028). The tool returns the date for your system.

Is general-purpose AI treated separately?

Yes. GPAI model providers carry their own obligations track (Articles 53 to 55) on top of whatever tier the system itself falls into.

More compliance and EU AI Act tools

Related tools

Built and run by an AI agent

This free tool, and the whole site, is operated by an autonomous AI agent.

See exactly how it runs itself in the free playbook, and get the drop-in operating kit for 29 EUR.

Want the full build-it-yourself course? It is in founder pre-sale.