Plantekia
Resources

Free resource · AI compliance

Does your AI meet the AI Act?

Answer a few questions and place your system in one of the four risk levels of the European AI Act. You will see the obligations and deadlines that apply to you. Indicative, in two minutes.

Why it matters

The AI Act clock is already running.

Regulation (EU) 2024/1689 applies in phases. From August 2026 the high-risk and transparency obligations come into force. Knowing which level each system falls into is the first step to neither arriving late nor overspending on unnecessary compliance.

The calculator

Classify your system in 2 minutes.

00 · Your role What is your relationship to the AI system?
01 · Prohibited practices Unacceptable risk (Art. 5)

If you tick any of these, the system is prohibited in the EU as of 2 February 2025.

02 · High risk High risk (Annex III / Annex I)

If you tick any of these, the system is high risk. Obligations apply from 2 August 2026 (or 2027 if it is embedded in a regulated product).

03 · Transparency Limited risk (Art. 50)

These do not raise the risk to high, but they create transparency obligations from 2 August 2026.

04 · General purpose GPAI models

Tick only if you build or fine-tune foundation models. It creates its own obligations from 2 August 2025.

This tool is indicative and does not constitute legal advice. The classification is based on Regulation (EU) 2024/1689 and may vary depending on the specific use case. For a binding assessment, consult a professional.

The four levels

From minimal risk to prohibited.

Unacceptable risk — Prohibited

In force since 2 February 2025

The system falls under a practice prohibited by Art. 5. It cannot be placed on the market or used in the European Union. The use case has to be redesigned or withdrawn.

High risk

Obligations apply from 2 August 2026 (Annex I: 2027)

The system affects safety or fundamental rights. It is legal, but subject to the most demanding compliance regime in the regulation before it goes to market.

Limited risk — Transparency

Transparency obligations from 2 August 2026

The system is not high risk, but it interacts with people or generates synthetic content. The law requires it to be clear that there is an AI behind it.

Minimal risk

No specific product obligations

Most systems (spam filters, recommenders, internal optimisation) land here. The AI Act imposes no product obligations, but it does encourage good practice.

Common questions

What we get asked most.

Is this calculator legally valid?

No. It is an indicative tool for placing your system within Regulation (EU) 2024/1689. The definitive classification depends on a full legal and technical analysis of each use case.

From when does the AI Act bind me?

In phases: prohibited practices and AI literacy apply from February 2025; general-purpose models from August 2025; Annex III high risk and transparency obligations from August 2026; and high risk embedded in regulated products from August 2027.

I run an SME — does it affect me the same way?

Yes, the regulation applies by type of system, not by company size. There are support measures for SMEs (sandboxes, simplified documentation), but the substantive obligations are the same if your system is high risk.

What is the difference between a provider and a deployer?

The provider develops or markets the system under its own brand and takes on the bulk of the compliance. The deployer uses it in its own operation and has lighter obligations: compliant use, human oversight and informing the people affected.

What are the penalties for non-compliance?

Up to €35M or 7% of annual worldwide turnover for prohibited practices, and up to €15M or 3% for breaching the rest of the obligations, with proportionate amounts for SMEs.

Let's start

Let's talk about your operation.

Tell us what costs more than it should. We come back with a concrete plan, realistic timelines and a clear yes or no.

contacto@plantekia.com