On 14 October 2026, the European Commission’s harmonised standards page recorded Official Journal publication of Commission Implementing Decision (EU) 2026/2274, adopted on 12 October 2026, on harmonised standards for high-risk AI systems under Regulation (EU) 2024/1689, the AI Act. That sounds procedural, but it gives the cited standards a defined legal role. Article 40 says high-risk AI systems or general-purpose AI models that comply with harmonised standards whose references are published in the Official Journal are presumed to meet the relevant AI Act requirements or obligations, to the extent those standards cover them. Harmonised standards page Article 40 page
What the public record does not show is almost as important. The Commission page names the implementing decision and its Official Journal citation, but it does not reproduce the referenced standards or their titles, and tells readers to contact the European Standardisation Organisations for the content and availability of European standards. The AI Act service-desk page explains the presumption-of-conformity mechanism, but also says its text is based on the EUR-Lex consolidated version as at 27 July 2026 and that its summaries are not legally binding. So the retrieved pages support a narrow conclusion: a legal route now exists, but they do not establish exactly which technical texts were cited or how far each one maps onto AI Act duties. Harmonised standards page Article 40 page
That matters because the main high-risk compliance dates are still ahead. The Commission’s enforcement overview says the rules for high-risk AI systems listed in Annex III apply from 2 December 2027, while the rules for high-risk AI systems embedded into regulated products apply from 2 August 2028. The same page says enforcement powers of the AI Office and national competent authorities apply from 2 August 2026 for provisions that have already become enforceable, but it also warns that the page is an information overview only and does not replace the legal text. Read against that timeline, the 14 October 2026 citation looks less like evidence of immediate market certification and more like a compliance route being put in place before the major Annex III deadline. Enforcement overview
The institutional map also matters. The Commission’s broader AI policy page says the European AI Office was established in June 2024 and has a role in implementation and enforcement of the AI Act in collaboration with Member States. The enforcement page adds that enforcement is shared between the AI Office, the European Data Protection Supervisor and national competent authorities, depending on the system and the user. That means an Official Journal citation is not the same thing as a finding by an enforcement authority, and the retrieved pages do not identify conformity assessments, certificates, notified bodies or market uptake. AI policy overview Enforcement overview
For vendors and buyers, the immediate change is documentary. A claim that a system follows European AI standards now has a more precise follow-up question: does that claim point to a standard whose reference appears in the Official Journal under the 12 October decision, or only to a draft, committee discussion or vendor interpretation? Because the Commission page does not list the referenced standards, procurement and compliance teams still need the decision citation and the underlying standards texts before they can judge whether a presumption-of-conformity claim is specific or vague. The same caution applies to third-party assessment claims, because the retrieved pages do not show who is assessing what. Harmonised standards page Article 40 page Enforcement overview
The contrast with the Commission’s other AI work helps separate legal effect from policy activity. On 30 September 2026, more than 90 signatories to the Code of Practice on Transparency of AI-generated Content set up two task forces, with meeting minutes still to be published soon and additional working groups only planned for October-November 2026 and January 2027. The Commission’s broader AI page also points to the April 2025 AI Continent action plan and the Apply AI Strategy. Those may shape practice, but they are not the same category of act as an Official Journal reference under Article 40. The useful distinction for readers is simple: some AI governance pages describe plans and forums, while the 14 October 2026 citation creates a defined legal presumption, even if the public record retrieved here leaves much of the practical detail unresolved. Task forces page AI policy overview Article 40 page
Read our editorial standards or send a correction to the editorial desk.
