Tekst instrumentu jest prowadzony po angielsku. Język interfejsu: Polski. English entry
EU AI Act, Article 14
Regulation (EU) 2024/1689 of the European Parliament and of the Council laying down harmonised rules on artificial intelligence, Article 14 (Human oversight)
Podsumowanie
Requires high-risk AI systems to be designed so an overseeing person can intervene in operation or interrupt the system through a stop button or similar procedure that allows it to come to a halt in a safe state. Application deferred to 2027-12-02.
Zakres
High-risk AI systems within the meaning of the AI Act.
Wymaga
- human-oversight
- safe-state
Daty
introduced: — | adopted: 2024-06-13 | inForce: — | appliesFrom: 2027-12-02
Źródła
- EUR-Lex, Publications Office of the European Union — 2026-10-05 — primary — https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX%3A02024R1689-20260727
- Council of the European Union — 2026-06-29 — primary — https://www.consilium.europa.eu/en/press/press-releases/2026/06/29/artificial-intelligence-council-gives-final-green-light-to-simplify-and-streamline-rules/
- European Commission, DG CONNECT (AI Act Service Desk) — 2026-09-25 — primary — https://ai-act-service-desk.ec.europa.eu/en/ai-act/article-14
Uwagi
The operative provision
Article 14(4)(e), quoted from the consolidated text of Regulation (EU) 2024/1689 as at 2026-07-27 in the Official Journal:
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.
Why regulates is recorded as both
Article 14(3) splits the oversight measures between those “built … into the high-risk AI system by the provider” and those “appropriate to be implemented by the deployer”. Article 14(4) then requires the system to be “provided to the deployer in such a way that natural persons to whom human oversight is assigned are enabled” to do the things listed, 14(4)(e) among them.
The provider builds the interruption capability. The deployer exercises it. The duty is
split in the Regulation’s own drafting, and the entry records both rather than choosing
one side of a text that chooses neither.
Amendment and application
Regulation (EU) 2026/1744 of 8 July 2026 was published in the Official Journal on
2026-07-24 and amends Article 113 of the AI Act. High-risk obligations apply from
2027-12-02 under Article 6(2) and Annex III, and from 2028-08-02 under Article 6(1) and
Annex I. The schema carries a single appliesFrom; it holds the Annex III date of
2027-12-02, and the Annex I date of 2028-08-02 is recorded here.
Article 14 itself is not amended. That has now been confirmed twice: from the amending regulation on 2026-09-28 and from the consolidation on 2026-10-05.
One provenance note: the retrieved consolidation truncated before Article 113, so the 2027-12-02 date rests on the text of Regulation (EU) 2026/1744 and not on the consolidation itself.