All terms

    High-Risk AI System (EU AI Act)

    AI systems captured by Article 6 of the EU AI Act because they are safety components of regulated products or listed in Annex III.

    Reviewed by Christian Espinosa, Founder, Blue Goat CyberLast reviewed July 25, 2026

    Definition

    Under Article 6 of Regulation (EU) 2024/1689, an AI system is high-risk if it is a safety component of, or itself is, a product covered by Union harmonisation legislation listed in Annex I (which includes medical devices under the MDR and IVDR) and requires third-party conformity assessment, or if it is listed in Annex III. Nearly all AI-enabled Class IIa and above medical devices fall into the first branch.
    What the regulation says
    Article 6(1) states an AI system is high-risk where it is intended to be used as a safety component of a product, or is itself a product, covered by Annex I Union harmonisation legislation and is required to undergo a third-party conformity assessment.

    What this means in practice

    For MedTech, this means the AI Act does not create a separate CE mark: obligations layer on top of the MDR/IVDR conformity assessment through the notified body. Manufacturers must satisfy Articles 8 to 15 (risk management, data governance, technical documentation, record-keeping, transparency, human oversight, accuracy, robustness, and cybersecurity) in addition to the underlying device regulation.

    Examples

    • An AI system used in a medical imaging device to detect subtle anomalies for cancer screening, where inaccurate detection could lead to missed diagnoses or unnecessary procedures.
    • An AI-powered algorithm embedded in a robotic surgical system that assists with precision movements, classifying as high-risk due to its direct safety component role in a Class IIb medical device.
    • A clinical decision support system that provides diagnostic or therapeutic recommendations based on patient data, if it directly impacts treatment outcomes and is subject to third-party conformity assessment under the MDR.
    Common pitfalls
    • •Assuming Class I MDR devices with AI escape the AI Act. Many still do, but only where no third-party conformity assessment is required.
    • •Treating the AI Act as a replacement for MDR Annex XVI clinical evaluation. It is additive, not substitutive.
    • •Confusing Annex III (standalone high-risk use cases) with the MedTech route through Annex I.

    Frequently asked questions

    Any AI-enabled medical device that requires notified-body involvement under MDR or IVDR is high-risk under the AI Act. Self-declared Class I MDR software is not automatically high-risk.
    Grouped by theme
    Cited by

    Where this term appears across MedTech Terms.

    Latest in MedTech

    Sources

    3 sources

    Every citation below opens the original document. Each is graded against our source-tier hierarchy so you can see what rests on binding law versus commentary.

    Tier 1Binding law and standards· 1Tier 2Regulator guidance and consensus· 2
    Link health: 2 verified 1 unchecked· last checked 2026-06-20
    EUR-Lex·1European Commission·1IMDRF·1
    1. 1
      Regulation (EU) 2024/1689, Article 6
      Tier 1 Verified
      EUR-Lexeur-lex.europa.eu
    2. 2
      Commission Q&A on the AI Act
      Tier 2 Unchecked
      European Commissiondigital-strategy.ec.europa.eu
    3. 3
      IMDRF Documents
      Tier 2 Verified
      IMDRFimdrf.org

    Inline markers like [1] jump to the matching reference above.