Correction: the high-risk deadlines were extended
An earlier version of this article incorrectly stated that the EU AI Act's high-risk provisions became fully enforceable in August 2026. The EU's AI Omnibus entered into force on 27 July 2026 and extended the application dates. Rules for the Annex III high-risk use cases apply from 2 December 2027, while high-risk AI systems embedded in regulated physical products under Annex I apply from 2 August 2028.
The broader AI Act did enter a new enforcement phase on 2 August 2026. That date brought additional provisions into application, including transparency requirements for certain AI-generated and manipulated content. It should not be confused with the later deadlines for the high-risk system obligations.
What still counts as high-risk
The change is to the timetable, not the underlying idea of a high-risk category. Annex III covers specified uses in areas including biometrics, critical infrastructure, education, employment, migration, asylum and border control. Product-related high-risk systems are covered through the separate Annex I route.
UK companies should therefore keep mapping whether products or deployments fall into those categories even though the compliance dates are later than previously expected. The additional time is intended to allow standards, guidance and other support measures to be in place before the obligations apply.
Why UK companies can still be in scope
Being headquartered in the UK does not by itself remove a company from the AI Act. Providers and deployers need to assess the Act's territorial scope when AI systems are placed on the EU market, put into service in the EU or otherwise fall within the regulation's cross-border provisions.
For UK vendors with European customers, the practical task is to identify which obligations apply now and which are tied to the December 2027 or August 2028 high-risk dates. Contracts, technical documentation, risk management and procurement processes should reflect that distinction rather than treating the entire Act as one deadline.
What businesses should do with the extra time
The extension should be treated as implementation time rather than a reason to stop preparation. Companies can use it to inventory AI systems, confirm classification, document data and model governance, establish human oversight where required and track the final Commission guidance and harmonised standards.
The key compliance mistake now would be the opposite of the one in the earlier version of this article: assuming that because the high-risk dates moved, nothing in the AI Act applies yet. Different provisions have different application dates, and the Commission began enforcing the rules that were already applicable from 2 August 2026.
Frequently asked questions
When do the EU AI Act high-risk rules apply?
Annex III high-risk rules apply from 2 December 2027. High-risk AI embedded in regulated physical products under Annex I applies from 2 August 2028.
Did the AI Act start being enforced in August 2026?
Yes, for provisions whose application date had arrived, including new transparency requirements. The high-risk system obligations have later application dates after the AI Omnibus extension.
Can the EU AI Act apply to UK companies?
Yes. UK companies need to assess territorial scope where they place AI systems on the EU market, put them into service in the EU or otherwise meet the Act's cross-border scope rules.