Skip to main content
European Commission logo
AI Act Service Desk

AI Act Explorer

Recitals
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180
Chapter I: General Provisions
Chapter II: Prohibited AI Practices
Chapter III: High-Risk AI Systems

Section 1: Classification of AI Systems as High-Risk

Section 2: Requirements for High-Risk AI Systems

Section 3: Obligations of Providers and Deployers of High-Risk AI Systems and Other Parties

Section 4: Notifying Authorities and Notified Bodies

Section 5: Standards, Conformity Assessment, Certificates, Registration

Chapter IV: Transparency Obligations for Providers and Deployers of Certain AI Systems
Chapter V: General-Purpose AI Models

Section 1: Classification Rules

Section 2: Obligations for Providers of General-Purpose AI Models

Section 3: Obligations of Providers of General-Purpose AI Models with Systemic Risk

Section 4: Codes of Practice

Chapter VI: Measures in Support of Innovation
Chapter VII: Governance

Section 1: Governance at Union Level

Section 2: National competent authorities

Chapter VIII: EU Database for High-Risk AI Systems
Chapter IX: Post-Market Monitoring, Information Sharing and Market Surveillance

Section 1: Post-Market Monitoring

Section 2: Sharing of Information on Serious Incidents

Section 3: Enforcement

Section 4: Remedies

Section 5: Supervision, Investigation, Enforcement and Monitoring in Respect of Providers of General-Purpose AI Models

Chapter X: Codes of Conduct and Guidelines
Chapter XI: Delegation of Power and Committee Procedure
Chapter XII: Penalties
Chapter XIII: Final Provisions
Annexes
Annex I
Annex II
Annex III
Annex IV
Annex V
Annex VI
Annex VII
Annex VIII
Annex IX
Annex X
Annex XI
Annex XII
Annex XIII
  • Chapter III: High-Risk AI Systems
  • Section 2: Requirements for High-Risk AI Systems

Article 15: Accuracy, robustness and cybersecurity

Summary

High-risk AI systems must be designed to achieve an appropriate level of accuracy, robustness, and cybersecurity throughout their lifecycle. The levels of accuracy and the relevant accuracy metrics of high-risk AI systems shall be declared in the accompanying instructions of use. High-risk AI systems should be resilient to errors, inconsistencies and faults. Cybersecurity should be ensured, with measures to prevent cyber-attacks, including data and model poisoning, adversarial examples, and confidentiality breaches. Technical solutions should be tailored to the system's risks and context.

The summaries are meant to provide helpful explanation but are not legal binding.

1.   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.

2.   To address the technical aspects of how to measure the appropriate levels of accuracy and robustness set out in paragraph 1 and any other relevant performance metrics, the Commission shall, in cooperation with relevant stakeholders and organisations such as metrology and benchmarking authorities, encourage, as appropriate, the development of benchmarks and measurement methodologies.

3.   The levels of accuracy and the relevant accuracy metrics of high-risk AI systems shall be declared in the accompanying instructions of use.

4.   High-risk AI systems shall be as resilient as possible regarding errors, faults or inconsistencies that may occur within the system or the environment in which the system operates, in particular due to their interaction with natural persons or other systems. Technical and organisational measures shall be taken in this regard.

The robustness of high-risk AI systems may be achieved through technical redundancy solutions, which may include backup or fail-safe plans.

High-risk AI systems that continue to learn after being placed on the market or put into service shall be developed in such a way as to eliminate or reduce as far as possible the risk of possibly biased outputs influencing input for future operations (feedback loops), and as to ensure that any such feedback loops are duly addressed with appropriate mitigation measures.

5.   High-risk AI systems shall be resilient against attempts by unauthorised third parties to alter their use, outputs or performance by exploiting system vulnerabilities.

The technical solutions aiming to ensure the cybersecurity of high-risk AI systems shall be appropriate to the relevant circumstances and the risks.

The technical solutions to address AI specific vulnerabilities shall include, where appropriate, measures to prevent, detect, respond to, resolve and control for attacks trying to manipulate the training data set (data poisoning), or pre-trained components used in training (model poisoning), inputs designed to cause the AI model to make a mistake (adversarial examples or model evasion), confidentiality attacks or model flaws.

Relevant recitals