Docs

Appearance

Operationalize the EU AI Act in Modulos

Requirements for high-risk AI systems ​

Articles 8 to 15 of the EU AI Act (Regulation (EU) 2024/1689) are the requirements for high-risk AI systems (Chapter III Section 2). Providers of high-risk AI systems must make sure that their systems comply with these requirements (Article 16(a)). Articles 8 to 15 do not apply to Article 6(1) systems related to Annex I Section B products. See Annex I Sections A and B.

This page also gives Article 4a on bias detection and correction. To find if an AI system is high-risk, see Is the AI system high-risk?.

Article 8 — compliance with the requirements ​

Article 8(1) of the EU AI Act gives the general rule for the Chapter III Section 2 requirements for high-risk AI systems:

High-risk AI systems shall comply with the requirements laid down in this Section, taking into account their intended purpose as well as the generally acknowledged state of the art on AI and AI-related technologies. The risk management system referred to in Article 9 shall be taken into account when ensuring compliance with those requirements.

In Article 8(1), each Section 2 requirement applies to a system in relation to its intended purpose and to the "generally acknowledged state of the art".

Article 8(2) applies to a product that contains an AI system to which the EU AI Act and Annex I Section A legislation apply. The provider is responsible for full compliance of the product with all applicable Union harmonization legislation. The provider can integrate, "as appropriate", the necessary testing and reporting processes, information and documentation into the procedures of the Section A legislation. This integration is an alternative that the provider can select. For the related Section A rules, see Annex I Section A products.

Article 9 — risk-management system ​

Article 9 of the EU AI Act tells providers of high-risk AI systems to have a risk management system. Article 9(1) gives the obligation:

A risk management system shall be established, implemented, documented and maintained in relation to high-risk AI systems.

Article 9(2) gives the system and its steps:

The risk management system shall be understood as a continuous iterative process planned and run throughout the entire lifecycle of a high-risk AI system, requiring regular systematic review and updating. It shall comprise the following steps:

(a) the identification and analysis of the known and the reasonably foreseeable risks that the high-risk AI system can pose to health, safety or fundamental rights when the high-risk AI system is used in accordance with its intended purpose;

(b) the estimation and evaluation of the risks that may emerge when the high-risk AI system is used in accordance with its intended purpose, and under conditions of reasonably foreseeable misuse;

(c) the evaluation of other risks possibly arising, based on the analysis of data gathered from the post-market monitoring system referred to in Article 72;

(d) the adoption of appropriate and targeted risk management measures designed to address the risks identified pursuant to point (a).

The risk management system is a continuous procedure for the full lifecycle of the system. It is not one document before the system is placed on the market. Point (c) uses data from the post-market monitoring system in Article 72 (see Article 72).

Article 9 also gives these rules:

  • Risks in scope (Article 9(3)). The risks are only those "which may be reasonably mitigated or eliminated through the development or design of the high-risk AI system, or the provision of adequate technical information".

  • Combined requirements (Article 9(4)). The risk management measures give "due consideration to the effects and possible interaction resulting from the combined application of the requirements" of Section 2.

  • Residual risk (Article 9(5)). The "relevant residual risk associated with each hazard" and the overall residual risk of the system must be "judged to be acceptable". To select the measures, the provider must do these steps:

    1. Remove or decrease the risks "in as far as technically feasible through adequate design and development of the high-risk AI system".
    2. Where "appropriate", apply "adequate mitigation and control measures addressing risks that cannot be eliminated".
    3. Give the information in Article 13 and, where "appropriate", training to deployers.

    For risks related to use, "due consideration" goes to the technical knowledge, experience, education and expected training of the deployer, and to "the presumable context in which the system is intended to be used".

  • Testing (Article 9(6) to (8)). The provider tests the system to identify the most "appropriate and targeted" risk management measures. The tests make sure that the system performs consistently for its intended purpose and complies with Section 2. Testing can include testing in real-world conditions in Article 60. Testing occurs, "as appropriate", at "any time throughout the development process" and, in all cases, before the system is placed on the market or put into service. Testing uses "prior defined metrics and probabilistic thresholds that are appropriate to the intended purpose".

  • Vulnerable groups (Article 9(9)). The provider gives "consideration to whether in view of its intended purpose the high-risk AI system is likely to have an adverse impact on persons under the age of 18 and, as appropriate, other vulnerable groups".

  • Other Union law (Article 9(10)). Other Union law can give a provider requirements for internal risk management processes. The provider can then make Article 9(1) to (9) "part of, or combined with" those procedures.

Article 10 — data and data governance ​

Article 10(1) of the EU AI Act, as amended, gives the data rule for high-risk AI systems that use the training of AI models:

High-risk AI systems which make use of techniques involving the training of AI models with data shall be developed on the basis of training, validation and testing data sets that meet the quality criteria referred to in paragraphs 2, 3 and 4 of this Article and in Article 4a(1) whenever such data sets are used.

Article 10(2) gives the data governance and management practices for training, validation and testing data sets:

Training, validation and testing data sets shall be subject to data governance and management practices appropriate for the intended purpose of the high-risk AI system. Those practices shall concern in particular:

(a) the relevant design choices;

(b) data collection processes and the origin of data, and in the case of personal data, the original purpose of the data collection;

(c) relevant data-preparation processing operations, such as annotation, labelling, cleaning, updating, enrichment and aggregation;

(d) the formulation of assumptions, in particular with respect to the information that the data are supposed to measure and represent;

(e) an assessment of the availability, quantity and suitability of the data sets that are needed;

(f) examination in view of possible biases that are likely to affect the health and safety of persons, have a negative impact on fundamental rights or lead to discrimination prohibited under Union law, especially where data outputs influence inputs for future operations;

(g) appropriate measures to detect, prevent and mitigate possible biases identified according to point (f);

(h) the identification of relevant data gaps or shortcomings that prevent compliance with this Regulation, and how those gaps and shortcomings can be addressed.

Article 10(3) tells providers that the data sets must be "relevant, sufficiently representative, and to the best extent possible, free of errors and complete in view of the intended purpose". The data sets must also have "the appropriate statistical properties", where applicable for the persons or groups on whom the system is intended for use. Each data set, or a combination of data sets, can have these characteristics.

Article 10(4) tells providers that the data sets "shall take into account, to the extent required by the intended purpose, the characteristics or elements that are particular to the specific geographical, contextual, behavioural or functional setting" of use.

Regulation (EU) 2026/1744 removed Article 10(5). Article 4a has the rules for the processing of special categories of personal data for bias detection and correction (see Article 4a).

Article 10(6) applies to high-risk AI systems that do not use the training of AI models. For these systems, Article 10(2) to (4) and Article 4a(1) apply only to the testing data sets.

Article 11 — technical documentation ​

Article 11(1) of the EU AI Act tells providers to prepare the technical documentation of a high-risk AI system before the system is placed on the market or put into service. Also, the provider must keep the documentation up to date. As a result, the documentation is a current record for the life of the system. Article 11(1) continues:

That technical documentation shall be drawn up in such a way as to demonstrate that the high-risk AI system complies with the requirements set out in this Section and to provide national competent authorities and notified bodies with the necessary information in a clear and comprehensive form to assess the compliance of the AI system with those requirements. It shall contain, at a minimum, the elements set out in Annex IV. […]

Annex IV gives the minimum content of the technical documentation, "as applicable to the relevant AI system":

  1. A general description of the AI system. Examples are its intended purpose, versions, forms of supply, hardware and instructions for use.
  2. A "detailed description of the elements of the AI system and of the process for its development". Examples are the "design specifications of the system", the system architecture, data requirements, human oversight assessment, pre-determined changes, and validation and testing procedures. Where "relevant", data requirements include datasheets about training methodologies, techniques and training data sets used (Annex IV point 2(d)).
  3. Information about the monitoring, functioning and control of the system, for example its capabilities and limitations in performance.
  4. A description of the appropriateness of the performance metrics.
  5. A "detailed description of the risk management system in accordance with Article 9".
  6. A description of "relevant changes made by the provider to the system through its lifecycle".
  7. The "list of the harmonised standards applied in full or in part" with "references" published in the Official Journal of the European Union. Where the provider applied no such standards, a detailed description of solutions adopted for compliance with Chapter III Section 2. This includes a "list of other relevant standards and technical specifications applied".
  8. A copy of the EU declaration of conformity in Article 47.
  9. A "detailed description of the system in place to evaluate the AI system performance in the post-market phase in accordance with Article 72, including the post-market monitoring plan referred to in Article 72(3)".

Article 11(2) applies to a high-risk AI system related to a product in Annex I Section A legislation. For such a system, the provider prepares "a single set of technical documentation". The set contains the information in Article 11(1) and the information that the Section A legislation makes necessary.

For the simplified technical documentation of SMEs and SMCs, see SMEs and SMCs.

Article 12 — record-keeping (logging) ​

Article 12(1) of the EU AI Act gives the logging requirement for high-risk AI systems:

High-risk AI systems shall technically allow for the automatic recording of events (logs) over the lifetime of the system.

Article 12(2) gives the events that the logging capabilities must record:

In order to ensure a level of traceability of the functioning of a high-risk AI system that is appropriate to the intended purpose of the system, logging capabilities shall enable the recording of events relevant for:

(a) identifying situations that may result in the high-risk AI system presenting a risk within the meaning of Article 79(1) or in a substantial modification;

(b) facilitating the post-market monitoring referred to in Article 72; and

(c) monitoring the operation of high-risk AI systems referred to in Article 26(5).

Article 12(3) gives the minimum logging capabilities for remote biometric identification systems:

For high-risk AI systems referred to in point 1 (a), of Annex III, the logging capabilities shall provide, at a minimum:

(a) recording of the period of each use of the system (start date and time and end date and time of each use);

(b) the reference database against which input data has been checked by the system;

(c) the input data for which the search has led to a match;

(d) the identification of the natural persons involved in the verification of the results, as referred to in Article 14(5).

Article 19 tells providers to keep the logs that their high-risk AI systems make automatically, when and while those logs are in their control, for the period that Article 19 gives. See provider obligations in Article 16.

Article 13 — transparency and provision of information to deployers ​

Article 13(1) of the EU AI Act gives the transparency requirement for high-risk AI systems:

High-risk AI systems shall be designed and developed in such a way as to ensure that their operation is sufficiently transparent to enable deployers to interpret a system’s output and use it appropriately. An appropriate type and degree of transparency shall be ensured with a view to achieving compliance with the relevant obligations of the provider and deployer set out in Section 3.

Article 13(2) tells providers to supply instructions for use with each high-risk AI system, "in an appropriate digital format or otherwise". The information must be "concise, complete, correct and clear" and "relevant, accessible and comprehensible to deployers". Article 13(3) gives the minimum content:

The instructions for use shall contain at least the following information:

(a) the identity and the contact details of the provider and, where applicable, of its authorised representative;

(b) the characteristics, capabilities and limitations of performance of the high-risk AI system, including:

(i) its intended purpose;

(ii) the level of accuracy, including its metrics, robustness and cybersecurity referred to in Article 15 against which the high-risk AI system has been tested and validated and which can be expected, and any known and foreseeable circumstances that may have an impact on that expected level of accuracy, robustness and cybersecurity;

(iii) any known or foreseeable circumstance, related to the use of the high-risk AI system in accordance with its intended purpose or under conditions of reasonably foreseeable misuse, which may lead to risks to the health and safety or fundamental rights referred to in Article 9(2);

(iv) where applicable, the technical capabilities and characteristics of the high-risk AI system to provide information that is relevant to explain its output;

(v) when appropriate, its performance regarding specific persons or groups of persons on which the system is intended to be used;

(vi) when appropriate, specifications for the input data, or any other relevant information in terms of the training, validation and testing data sets used, taking into account the intended purpose of the high-risk AI system;

(vii) where applicable, information to enable deployers to interpret the output of the high-risk AI system and use it appropriately;

(c) the changes to the high-risk AI system and its performance which have been pre-determined by the provider at the moment of the initial conformity assessment, if any;

(d) the human oversight measures referred to in Article 14, including the technical measures put in place to facilitate the interpretation of the outputs of the high-risk AI systems by the deployers;

(e) the computational and hardware resources needed, the expected lifetime of the high-risk AI system and any necessary maintenance and care measures, including their frequency, to ensure the proper functioning of that AI system, including as regards software updates;

(f) where relevant, a description of the mechanisms included within the high-risk AI system that allows deployers to properly collect, store and interpret the logs in accordance with Article 12.

Article 14 — human oversight ​

Article 14(1) of the EU AI Act gives the human oversight requirement for high-risk AI systems:

High-risk AI systems shall be designed and developed in such a way, including with appropriate human-machine interface tools, that they can be effectively overseen by natural persons during the period in which they are in use.

Human oversight "shall aim to prevent or minimise the risks to health, safety or fundamental rights" in use for the intended purpose or with reasonably foreseeable misuse. This applies "in particular where such risks persist despite the application of other requirements" of Section 2 (Article 14(2)).

Article 14(3) tells providers that the oversight measures must be "commensurate with the risks, level of autonomy and context of use". The measures must use one or more of these types:

  • Measures that the provider identifies and puts into the system, "when technically feasible", before the system is placed on the market or put into service (point (a)).
  • Measures that the provider identifies before the system is placed on the market or put into service, and that are "appropriate to be implemented by the deployer" (point (b)).

Article 14(4) tells providers how to supply the system to the deployer:

For the purpose of implementing paragraphs 1, 2 and 3, the high-risk AI system shall be provided to the deployer in such a way that natural persons to whom human oversight is assigned are enabled, as appropriate and proportionate:

(a) to properly understand the relevant capacities and limitations of the high-risk AI system and be able to duly monitor its operation, including in view of detecting and addressing anomalies, dysfunctions and unexpected performance;

(b) to remain aware of the possible tendency of automatically relying or over-relying on the output produced by a high-risk AI system (automation bias), in particular for high-risk AI systems used to provide information or recommendations for decisions to be taken by natural persons;

(c) to correctly interpret the high-risk AI system’s output, taking into account, for example, the interpretation tools and methods available;

(d) to decide, in any particular situation, not to use the high-risk AI system or to otherwise disregard, override or reverse the output of the high-risk AI system;

(e) to intervene in the operation of the high-risk AI system or interrupt the system through a ‘stop’ button or a similar procedure that allows the system to come to a halt in a safe state.

Article 14(5) applies to remote biometric identification systems in Annex III, point 1(a). The Article 14(3) measures must also make sure that the deployer acts or decides on an identification from the system only if two or more natural persons independently verified and confirmed it. These persons must have "the necessary competence, training and authority". This rule does not apply to high-risk AI systems for law enforcement, migration, border control or asylum, where Union or national law considers it "disproportionate".

Deployers must give the human oversight task to natural persons who have "the necessary competence, training and authority, as well as the necessary support" (Article 26(2)). See deployer obligations in Article 26.

Article 15 — accuracy, robustness and cybersecurity ​

Article 15(1) of the EU AI Act gives the requirement for accuracy, robustness and cybersecurity of high-risk AI systems:

High-risk AI systems shall be designed and developed in such a way that they achieve an appropriate level of accuracy, robustness, and cybersecurity, and that they perform consistently in those respects throughout their lifecycle.

Article 15 also gives these rules:

  • Benchmarks (Article 15(2)). The Commission must encourage, "as appropriate", the development of benchmarks and measurement methods. It does this in cooperation with stakeholders and organizations, for example metrology and benchmarking authorities.
  • Declared accuracy (Article 15(3)). The instructions for use must declare the levels of accuracy and the related accuracy metrics of the system.
  • Robustness (Article 15(4)). High-risk AI systems must be "as resilient as possible regarding errors, faults or inconsistencies" in the system or its environment. Technical and organizational measures are necessary for this. Robustness can use technical redundancy solutions, for example "backup or fail-safe plans".
  • Feedback loops (Article 15(4)). Some systems "continue to learn after being placed on the market or put into service". These systems must be developed to remove or decrease "as far as possible" the risk of "possibly biased outputs influencing input for future operations (feedback loops)". Such feedback loops must have "appropriate mitigation measures".
  • Cybersecurity (Article 15(5)). High-risk AI systems must be resilient against "attempts by unauthorised third parties to alter their use, outputs or performance by exploiting system vulnerabilities". The technical solutions must be "appropriate to the relevant circumstances and the risks".
  • AI-specific vulnerabilities (Article 15(5)). The solutions include, where "appropriate", "measures to prevent, detect, respond to, resolve and control for attacks". Examples are data poisoning, model poisoning, adversarial examples or model evasion, confidentiality attacks and model flaws.

Article 4a — bias detection and correction ​

Article 4a of the EU AI Act permits providers and deployers to process special categories of personal data exceptionally for bias detection and correction on its conditions. Regulation (EU) 2026/1744 entered into force on 27 July 2026. It added Article 4a in Chapter I and removed Article 10(5). For the special categories, Article 3(37) refers to Regulation (EU) 2016/679, Directive (EU) 2016/680 and Regulation (EU) 2018/1725.

Article 4a(1) applies to providers of high-risk AI systems:

To the extent strictly necessary to ensure bias detection and correction in relation to high-risk AI systems in accordance with Article 10(2), points (f) and (g), of this Regulation, providers of such systems may exceptionally process special categories of personal data, subject to appropriate safeguards for the fundamental rights and freedoms of natural persons. In addition to the provisions set out in Regulations (EU) 2016/679 and (EU) 2018/1725 and Directive (EU) 2016/680, as applicable, all the following conditions shall be met in order for such processing to occur:

(a) the bias detection and correction cannot be effectively fulfilled by processing other data, including synthetic or anonymised data;

(b) the special categories of personal data are subject to technical limitations on the re-use of personal data, and state-of-the-art security and privacy-preserving measures, including pseudonymisation;

(c) the special categories of personal data are subject to measures to ensure that the personal data processed are secured and protected, subject to suitable safeguards, including strict controls and documentation of the access, to avoid misuse and to ensure that only authorised persons have access to those personal data with appropriate confidentiality obligations;

(d) the special categories of personal data are not transmitted, transferred or otherwise accessed by other parties;

(e) the special categories of personal data are deleted once the bias has been corrected or the personal data has reached the end of its retention period, whichever comes first; and

(f) the records of processing activities pursuant to Regulations (EU) 2016/679 and (EU) 2018/1725 and Directive (EU) 2016/680 include the reasons why the processing of special categories of personal data was strictly necessary to detect and correct biases, and why that objective could not be achieved by processing other data.

Article 4a(2) applies to providers and deployers of other AI systems and models, and to deployers of high-risk AI systems. They can "exceptionally process special categories of personal data" only if the two conditions in Article 4a(2) apply together:

  • The processing is "strictly necessary to ensure bias detection and correction in view of possible biases that are likely to affect the health and safety of persons, have a negative impact on fundamental rights or lead to discrimination prohibited pursuant to Union law, especially where data outputs influence inputs for future operations".
  • "all of the conditions and safeguards set out in paragraph 1 are applied".

Article 4a(2) "does not create any obligation to conduct such bias detection and correction".

Article 2(7) preserves Union data-protection law for personal data processed in connection with EU AI Act rights and obligations, without prejudice to Articles 4a and 59. Article 4a does not apply to Article 6(1) systems related to Annex I Section B products (Article 2(2)).

Modulos ​

Modulos has EU AI Act Frameworks at Organization level (OFF-1) and at AI System level (MFF-1). For the Requirements that map Articles 4a and 8 to 15, see Requirements and provisions. For the sequence of work that Modulos recommends, see Complete the work in sequence.

Sources ​

  • Regulation (EU) 2024/1689 (EU AI Act), consolidated version of 27 July 2026:
    • Articles 2(2), 2(7), 3(37) and 4a.
    • Articles 8 to 16, 19 and 26(2).
    • Annexes III and IV.
  • Regulation (EU) 2026/1744 of 8 July 2026 (Digital Omnibus on AI), Official Journal of 24 July 2026: Article 1(6) and (9) and Article 4.

Disclaimer

This page gives general information. It is not legal advice.