Indicative self-assessment

Check how your AI system may be classified under the EU AI Act

Answer a guided set of questions to receive a preliminary classification and a report you can take into legal and risk review.

Guided questions
Preliminary result
Email report after confirmation
Check my system

How it works

A starting point for a qualified classification review

1

Describe the system

Tell us about your AI system, its purpose, and how it operates. Our guided wizard walks you through the key questions from the EU AI Act.

2

See a preliminary result

The rule engine maps your answers to a possible risk tier and shows the decision path. It does not replace legal analysis.

3

Take it into review

Confirm your email to receive a report with the preliminary result, relevant references, and questions to verify.

Step 1 of 5: System
1
System
2
Prohibited
3
High-Risk
4
Context
5
Your Details

About Your AI System

Tell us about the AI system you want to classify under the EU AI Act.

What is the name of your AI system or project?*

Briefly describe what your AI system does*

What is the primary purpose of your AI system?*

Where will this AI system be deployed or made available?*

What is your role in relation to this AI system?*

What you'll get

A self-assessment report to support your next review

Preliminary tier

A possible risk tier based on the information you provide.

Decision path

The answers and rule branches that led to the preliminary result.

Relevant references

EU AI Act references associated with the result for further review.

Questions to verify

Points to confirm with qualified legal, risk, and compliance owners.

Frequently Asked Questions

The EU AI Act categorises AI systems into four risk tiers: Unacceptable (banned), High (strict requirements), Limited (transparency obligations), and Minimal (voluntary codes). Your obligations depend on which tier your system falls into.

The tool maps your answers through a deterministic rule path to produce a preliminary result. It can miss facts or legal nuance, so it is a starting point for qualified review, not a legal determination.

The report includes the preliminary classification, the decision path created from your answers, relevant regulatory references, and questions to verify with legal, risk, and compliance owners.

Your submission is handled under the Difinity Privacy Policy. Review that policy before submitting personal or confidential information.

Email confirmation (double opt-in) ensures we deliver your report to a verified business email address and complies with data protection best practices. You'll receive a preliminary classification immediately, and the full report after email confirmation.

You can run the classifier as many times as needed for different AI systems. Each submission generates a separate report tailored to that specific system.