EU AI Act FAQ
What is a high-risk AI system under the EU AI Act?
Quick answer
High-risk classification follows Article 6: certain product or safety-component systems linked to Annex I legislation, and the specified Annex III uses subject to relevant exceptions. Employment, health or finance as a sector label is not enough. The exact intended purpose and effect on people determine the assessment.
Examples and boundaries
Annex III includes recruitment and candidate evaluation, specified education decisions, natural-person creditworthiness and life or health insurance risk assessment and pricing. It does not classify every use of AI in those sectors as high-risk. Financial-fraud detection is expressly excluded from the credit-scoring category. Some narrow or preparatory Annex III tasks can qualify for the Article 6(3) exception if the statutory conditions are met; profiling in an Annex III use remains high-risk.
What follows from classification
Provider responsibilities include the applicable conformity process and supporting documentation. Deployer responsibilities include using the system as instructed, competent oversight, monitoring and relevant records. Fundamental-rights assessment and database registration have particular scopes, not a blanket requirement for every private customer. The amended timeline distinguishes Annex III rules from the Annex I product route. Preserve the classification rationale and revisit it if the purpose or system changes.
Reviewed 4 October 2026. Sources: Commission AI Act overview and timetable, classification, deployer duties, literacy guidance and AI Office of Ireland.
Acuity AI helps Irish organisations identify and manage high-risk AI obligations under the EU AI Act. See our EU AI Act compliance services.