AR-006ÉlevéFournisseurCorrectif
Procédures de repli en cas de défaillance du système
Les fournisseurs doivent concevoir et mettre en œuvre des procédures de repli qui s'activent lorsque le système d'IA ne peut pas fonctionner dans ses paramètres de conception, garantissant une dégradation sûre du service et une notification aux superviseurs humains.
Articles:Article 15(3)Article 14
Exemples de preuves
- Fallback procedure specification
- Graceful degradation design description
- Fallback activation test records
Normes
ISO 42001:2023 §8.4
Contrôles associés
- AR-001Accuracy Level Definition and CommitmentArticle 15(1) requires that high-risk AI systems are designed and developed to achieve an appropriate level of accuracy, robustness, and cybersecurity,…
- AR-002Accuracy Measurement and VerificationProviders must measure and verify the accuracy of the AI system against defined metrics using representative test sets, documenting the methodology,…
- AR-003Robustness Testing Against Input VariationsProviders must test the AI system's robustness against input variations that may arise in real-world deployment, including incomplete data, noisy inputs,…
- AR-004Adversarial Robustness TestingWhere relevant to the AI system's use case, providers must test robustness against adversarial inputs designed to cause misclassification, data poisoning…
- AR-005Cybersecurity Measures for AI SystemsArticle 15(5) requires that high-risk AI systems are resilient against unauthorised third-party attempts to alter their outputs, and providers must…
- AR-007Input Validation and Data Quality ChecksProviders must implement input validation mechanisms that verify the quality, format, and plausibility of data before it is processed by the AI system,…
Termes associés du règlement sur l’IA
- Human OversightMeasures built into high-risk AI systems enabling natural persons to understand, monitor, and — where necessary — override or shut down the system. Must be proportionate to the risks and must ensure that deployers can intervene in the system's output before it takes effect.
- AccuracyThe requirement that high-risk AI systems achieve an appropriate level of accuracy in relation to their intended purpose, as specified in the technical documentation. Providers must declare the level of accuracy in the instructions for use.
- RobustnessThe ability of a high-risk AI system to maintain its level of performance under adverse conditions — including technical limitations, adversarial inputs, errors, or unexpected situations — or within foreseeable operating conditions outside the intended purpose.
- CybersecurityThe requirement that high-risk AI systems are resilient against attempts by third parties to alter their use, behaviour, or performance in ways that could result in risks to health, safety, or fundamental rights, including protection against data poisoning, adversarial examples, and model evasion attacks.
Passez à la vitesse supérieure quand la conformité doit être
Le questionnaire gratuit fournit des signaux préliminaires. L'Évaluation Complète transforme les données réelles de vos systèmes en un dossier de décision gouverné : extraction, séparation des composants, preuves, surcouches nationales et dossier prêt à être audité.
Démarrer l'aperçu gratuit des risques