Conduct High-Risk AI Real-World Testing Only Under the Article 60 Conditions

Sub-obligations

These are independently assessable parts of the parent requirement.

1. Article 60(4)(a)-(c)

   Current

   Prepare, Submit and Obtain Approval for the Real-World Testing Plan and Complete Required Registration

   Tracker Networks Guidance

   Prepare and submit the real-world testing plan, obtain the required market-surveillance approval or applicable tacit approval, and register the test through the correct EU, secure or national database route with the required unique identifier.

   Official text

   (a) the provider or prospective provider has drawn up a real-world testing plan and submitted it to the market surveillance authority in the Member State where the testing in real world conditions is to be conducted; (b) the market surveillance authority in the Member State where the testing in real world conditions is to be conducted has approved the testing in real world conditions and the real-world testing plan; where the market surveillance authority has not provided an answer within 30 days, the testing in real world conditions and the real-world testing plan shall be understood to have been approved; where national law does not provide for a tacit approval, the testing in real world conditions shall remain subject to an authorisation; [Excerpt - see official source for complete provision]

   Excerpt stored at a complete legal-unit boundary. See the official source for the full provision.

   Suggested controls

   • CE Marking and AI Registration

     Apply required CE marking and complete applicable EU or national AI registration before market placement, service or qualifying use.

   • High-Risk AI Real-World Testing Governance

     Plan, approve, register, govern and close real-world testing of high-risk AI under Articles 60/60a.

   Related risks

   • AI Declaration, CE Marking or Registration Failure

     Required declarations, CE marking or EU/national database registrations may be missing, inaccurate, outdated or completed too late.

   • Unsafe or Non-Compliant Real-World AI Testing

     Real-world testing may proceed without required approval, registration, participant protection, oversight, incident handling or closure controls.

2. Article 60(4)(d)-(e)

   Current

   Maintain EU Establishment or Representation and Apply Required Cross-Border Data Safeguards

   Tracker Networks Guidance

   Be established in the Union or appoint an EU-established legal representative and transfer testing data to third countries only with appropriate applicable Union-law safeguards.

   Official text

   (d) the provider or prospective provider conducting the testing in real world conditions is established in the Union or has appointed a legal representative who is established in the Union; (e) data collected and processed for the purpose of the testing in real world conditions shall be transferred to third countries only provided that appropriate and applicable safeguards under Union law are implemented;

   Suggested controls

   • AI Value Chain Contracting and Technical Handover

     Define information, technical access, assistance and handover responsibilities across the high-risk AI value chain.

   • High-Risk AI Real-World Testing Governance

     Plan, approve, register, govern and close real-world testing of high-risk AI under Articles 60/60a.

   Related risks

   • AI Value Chain Compliance Failure

     Suppliers, importers, distributors or authorised representatives may not provide the information, controls or cooperation needed for high-risk AI compliance.

   • Unsafe or Non-Compliant Real-World AI Testing

     Real-world testing may proceed without required approval, registration, participant protection, oversight, incident handling or closure controls.

3. Article 60(4)(f)-(g)

   Current

   Limit Real-World Testing Duration and Protect Vulnerable Participants

   Tracker Networks Guidance

   Limit testing to the time necessary and no more than six months unless the permitted additional six-month extension is notified and justified, and appropriately protect participants vulnerable due to age or disability.

   Official text

   (f) the testing in real world conditions does not last longer than necessary to achieve its objectives and in any case not longer than six months, which may be extended for an additional period of six months, subject to prior notification by the provider or prospective provider to the market surveillance authority, accompanied by an explanation of the need for such an extension; (g) the subjects of the testing in real world conditions who are persons belonging to vulnerable groups due to their age or disability, are appropriately protected;

   Suggested controls

   • High-Risk AI Real-World Testing Governance

     Plan, approve, register, govern and close real-world testing of high-risk AI under Articles 60/60a.

   Related risks

   • AI-Related Fundamental Rights or Discrimination Harm

     AI use may create discriminatory, unfair or other material adverse impacts on individuals or groups.

   • Unsafe or Non-Compliant Real-World AI Testing

     Real-world testing may proceed without required approval, registration, participant protection, oversight, incident handling or closure controls.

4. Article 60(4)(h)

   Current

   Define Provider and Deployer Responsibilities for Joint Real-World Testing

   Tracker Networks Guidance

   Where testing is conducted with deployers or prospective deployers, provide relevant testing information and instructions and enter into an agreement defining roles and responsibilities for compliance.

   Official text

   (h) where a provider or prospective provider organises the testing in real world conditions in cooperation with one or more deployers or prospective deployers, the latter have been informed of all aspects of the testing that are relevant to their decision to participate, and given the relevant instructions for use of the AI system referred to in Article 13; the provider or prospective provider and the deployer or prospective deployer shall conclude an agreement specifying their roles and responsibilities with a view to ensuring compliance with the provisions for testing in real world conditions under this Regulation and under other applicable Union and national law;

   Suggested controls

   • AI Value Chain Contracting and Technical Handover

     Define information, technical access, assistance and handover responsibilities across the high-risk AI value chain.

   • High-Risk AI Real-World Testing Governance

     Plan, approve, register, govern and close real-world testing of high-risk AI under Articles 60/60a.

   Related risks

   • AI Value Chain Compliance Failure

     Suppliers, importers, distributors or authorised representatives may not provide the information, controls or cooperation needed for high-risk AI compliance.

   • Unsafe or Non-Compliant Real-World AI Testing

     Real-world testing may proceed without required approval, registration, participant protection, oversight, incident handling or closure controls.

5. Article 60(4)(i)-(k), 60(5)

   Current

   Protect Testing Subjects, Maintain Qualified Oversight and Ensure Outputs Can Be Reversed or Disregarded

   Tracker Networks Guidance

   Obtain required informed consent or apply the limited law-enforcement alternative, respect withdrawal and deletion rights, ensure testing is overseen by suitably qualified persons and ensure predictions, recommendations or decisions can be effectively reversed and disregarded.

   Official text

   (i) the subjects of the testing in real world conditions have given informed consent in accordance with Article 61, or in the case of law enforcement, where the seeking of informed consent would prevent the AI system from being tested, the testing itself and the outcome of the testing in the real world conditions shall not have any negative effect on the subjects, and their personal data shall be deleted after the test is performed; (j) the testing in real world conditions is effectively overseen by the provider or prospective provider, as well as by deployers or prospective deployers through persons who are suitably qualified in the relevant field and have the necessary capacity, training and authority to perform their tasks; (k) the predictions, recommendations or decisions of the AI system can be effectively reversed and disregarded.

   5. Any subjects of the testing in real world conditions, or their legally designated representative, as appropriate, may, without any resulting detriment and without having to provide any justification, withdraw from the testing at any time by revoking their informed consent and may request the immediate and permanent deletion of their personal data. The withdrawal of the informed consent shall not affect the activities already carried out.

   Suggested controls

   • Human Oversight Design and Enablement

     Design systems and deployment guidance so qualified humans can monitor, interpret, override and safely stop high-risk AI.

   • High-Risk AI Real-World Testing Governance

     Plan, approve, register, govern and close real-world testing of high-risk AI under Articles 60/60a.

   • Real-World Testing Informed Consent

     Obtain and document legally sufficient informed consent for Article 60 real-world testing.

   Related risks

   • Ineffective Human Oversight of AI

     Human oversight may be nominal or ineffective because people lack authority, competence, information or usable intervention mechanisms.

   • Unsafe or Non-Compliant Real-World AI Testing

     Real-world testing may proceed without required approval, registration, participant protection, oversight, incident handling or closure controls.

6. Article 60(7)

   Current

   Mitigate Serious Incidents During Real-World Testing and Maintain a Recall Procedure

   Tracker Networks Guidance

   Report serious incidents under Article 73, take immediate mitigation measures, suspend or terminate testing if mitigation cannot be achieved and maintain a procedure for prompt recall of the AI system following termination.

   Official text

   7. Any serious incident identified in the course of the testing in real world conditions shall be reported to the national market surveillance authority in accordance with Article 73. The provider or prospective provider shall adopt immediate mitigation measures or, failing that, shall suspend the testing in real world conditions until such mitigation takes place, or otherwise terminate it. The provider or prospective provider shall establish a procedure for the prompt recall of the AI system upon such termination of the testing in real world conditions.

   Suggested controls

   • AI Incident Escalation and Notification

     Detect, triage and escalate serious AI incidents and relevant risk conditions within required timelines.

   • High-Risk AI Serious Incident Reporting

     Detect, triage, report, investigate and preserve evidence for serious high-risk AI incidents within applicable deadlines.

   • High-Risk AI Real-World Testing Governance

     Plan, approve, register, govern and close real-world testing of high-risk AI under Articles 60/60a.

   Related risks

   • Serious AI Incident Reporting Failure

     A serious AI incident may not be reported, investigated or preserved within the required timeline.

   • Unsafe or Non-Compliant Real-World AI Testing

     Real-world testing may proceed without required approval, registration, participant protection, oversight, incident handling or closure controls.

7. Article 60(8)

   Current

   Notify the Market-Surveillance Authority of Testing Suspension, Termination and Final Outcomes

   Tracker Networks Guidance

   Notify the national market-surveillance authority where testing occurs of any suspension or termination and of final testing outcomes.

   Official text

   8. Providers or prospective providers shall notify the national market surveillance authority in the Member State where the testing in real world conditions is to be conducted of the suspension or termination of the testing in real world conditions and of the final outcomes.

   Suggested controls

   • Regulatory Information and Cooperation Process

     Respond to reasoned regulator requests with controlled, complete and retrievable compliance information and logs.

   • High-Risk AI Real-World Testing Governance

     Plan, approve, register, govern and close real-world testing of high-risk AI under Articles 60/60a.

   Related risks

   • Regulatory Cooperation Failure

     The organisation may be unable to provide complete, timely and understandable information, records or cooperation to competent authorities.

   • Unsafe or Non-Compliant Real-World AI Testing

     Real-world testing may proceed without required approval, registration, participant protection, oversight, incident handling or closure controls.