Free tool · 3 minutes · no registration
EU AI Act risk classifier
Answer a few questions about one AI system and find out its risk category under the EU AI Act, the obligations for your role and the dates that apply. Nothing you enter leaves your browser.
How the EU AI Act classifies AI systems
Regulation (EU) 2024/1689, as amended by Regulation (EU) 2026/1744 (Digital Omnibus on AI), follows a risk-based approach:
- Prohibited practices (Art. 5): manipulation, exploitation of vulnerabilities, social scoring, emotion recognition at work or in education and other practices. In force since 02/02/2025; new prohibitions on non-consensual sexual content from 02/12/2026.
- High-risk systems (Art. 6): safety components of regulated products (Annex I, from 02/08/2028) and systems used in the areas of Annex III (from 02/12/2027), with obligations for providers, deployers, importers and distributors.
- Transparency obligations (Art. 50): chatbots, synthetic content, emotion recognition and deepfakes. Applicable since 02/08/2026.
- General-purpose AI models (Art. 51-55): documentation, copyright policy and, for models with systemic risk, additional safeguards. Applicable since 02/08/2025.
- Minimal risk: no specific obligations for the system, apart from AI literacy (Art. 4).
Checked against the consolidated text of 27/07/2026. Sources: Regulation (EU) 2024/1689, Regulation (EU) 2026/1744. Preliminary orientation, not legal advice.