Answer the questions below to determine the risk classification of your AI system under the EU AI Act. This typically takes 3-5 minutes.
Select a template to pre-fill answers based on your AI system type. You can change any answer during the questionnaire.
Conversational AI for customer support. Typically limited risk due to Art. 50 transparency requirements.
Content, product, or service recommendations based on user data. Typically minimal risk.
CV screening, candidate filtering, or interview assessment. High risk under Annex III, 4(a).
Creditworthiness assessment or credit score calculation. High risk under Annex III, 5(b).
Automated review of user-generated content for policy violations. Typically minimal risk.
AI for evaluating emergency calls or triaging emergency patients (e.g., 112/911 dispatch prioritization, ER triage). High risk under Annex III, 5(d). Note: general clinical decision support may follow a different pathway (Annex I / medical device legislation).
Financial fraud detection and prevention. Explicitly exempt from Annex III, 5(b) high-risk when used solely for fraud detection.
Automated document scanning, text extraction, and data entry. Typically minimal risk.
Generative AI creating synthetic images, video, or audio. Limited risk with Art. 50 transparency obligations.
Equipment monitoring and failure prediction for industrial systems. Typically minimal risk.
Start with a blank questionnaire and answer each question manually.