The AI Act Omnibus Regulation entered into force in July 2026 following the Council of the European Union's final approval on 29 June 2026. The Regulation amends Regulation (EU) 2024/1689 (the AI Act) by deferring the application of Chapter III obligations for high-risk AI systems. The Commission first proposed the deferral in November 2025 under the Digital Omnibus legislative package, citing delays in harmonised standards development and the incomplete establishment of national competent authorities across member states.
The amended AI Act postpones the application dates for high-risk AI system obligations established by the transitional provisions of Regulation (EU) 2024/1689. For standalone high-risk AI systems listed in Annex III, covering areas including biometrics, critical infrastructure management, education, employment, migration, asylum, and border control, obligations now apply from 2 December 2027. For high-risk AI systems embedded in products subject to EU product safety legislation (Annex II systems), obligations apply from 2 August 2028. The Omnibus also extends the deadline for national competent authorities to establish AI regulatory sandboxes to 2 August 2027.
Developers, deployers, and importers of high-risk AI systems gain additional time to prepare conformity assessments, build quality management systems under Article 17, finalise technical documentation under Article 11, and register systems in the EU database under Article 71. Organisations that aligned compliance timelines to August 2026 should revise their internal programs. The deferral does not affect the transparency obligations under Article 50, which entered into application on 2 August 2026, nor the general-purpose AI model obligations enforceable from the same date.
The Commission stated the postponement was warranted by delays in harmonised standards preparation under the mandate issued to CEN/CENELEC and the incomplete readiness of national market surveillance infrastructure. Whether the Commission will issue alternative common specifications under Article 40(3) before 2 December 2027 or will rely on voluntary harmonised standards remains open. Providers of high-risk AI systems should monitor the Commission's standardisation workplan and guidance from the European AI Office for progress on technical specifications required for conformity assessment.
Licentium advises businesses on EU AI Act compliance across all risk categories, including classification under Annexes II and III, conformity assessment preparation, technical documentation review, and engagement with national competent authorities. Work we undertake includes EU AI Act advisory, high-risk AI system classification, GPAI model compliance, conformity assessment support, AI regulatory sandbox applications, and EU digital regulation advisory.