Blab AI
EU Council Approves Omnibus VII AI Regulation, Delays High-Risk AI Obligations
On 29 June 2026 the European Council formally approved a new regulation that will streamline and simplify the EU’s artificial‑intelligence (AI) framework. The measure is part of the Omnibus VII legislative package, which seeks to reduce administrative burdens across a range of EU policies.
The regulation postpones the entry into force of several high‑risk AI obligations that were originally scheduled for 2 August 2026. Under the new timetable, stand‑alone high‑risk AI systems will become subject to the AI Act on 2 December 2027, while high‑risk AI embedded in products will be regulated from 2 August 2028.
Key provisions of the new regulation include:
A prohibition on AI systems that generate non‑consensual sexual or intimate content, including child sexual abuse material (CSAM). The ban covers AI that creates nude images of real people or edits photographs to remove clothing and reveal intimate parts. The prohibition takes effect in December 2026. A postponement of the deadline for national authorities to establish AI regulatory sandboxes. The new deadline is 2 August 2027, one year later than the original 2 August 2026 target. A reduction of the grace period for AI providers to implement transparency solutions for artificially generated content. The period is cut from six months to three months, with the new deadline set for 2 December 2026. Clarification of the European AI Office’s supervisory role over general‑purpose AI models that are developed and deployed by the same provider. The Office’s competence is limited where national authorities retain jurisdiction, such as law enforcement, border management, judicial authorities, and financial institutions. A mechanism to resolve overlaps between the AI Act and sector‑specific legislation. In cases where sectoral law contains similar AI‑specific requirements—examples include medical devices, toys, lifts, and watercraft—the AI Act’s high‑risk provisions can be limited through implementing acts. Exemption of products covered by the machinery regulation from direct applicability of the AI Act. The Commission may adopt secondary legislation to add health and safety requirements for high‑risk AI systems that fall under the machinery regulation. * A new obligation for the Commission to issue guidance that helps operators of high‑risk AI systems, which are also subject to sectoral harmonisation legislation, comply with the AI Act while minimising administrative burden.
The regulation will be published in the EU’s Official Journal shortly. It will enter into force on the third day after publication.
The move follows a series of EU initiatives aimed at simplifying digital regulation. In October 2024 the European Council urged all EU institutions and member states to accelerate work on simplification, citing reports by Enrico Letta and Mario Draghi. The Budapest declaration of 8 November 2024 called for a “simplification revolution” to reduce regulatory and reporting burdens, especially for small and medium‑sized enterprises. Since February 2025 the Commission has released ten Omnibus packages covering sustainability, investment, agriculture, digitalisation, defence, chemicals, automotive, food safety, and more.
The high‑risk AI delay aligns with the broader Omnibus VII agenda, which keeps the core architecture of the AI Act unchanged while providing additional time for providers and regulators to prepare. The postponement does not alter the AI Act’s core prohibitions or transparency obligations for general‑purpose AI.
The regulation’s ban on non‑consensual sexual content and CSAM reflects the EU’s ongoing effort to address harmful uses of generative AI. The AI Office’s expanded oversight role and the new guidance obligation aim to balance innovation with consumer protection.
In the coming weeks, the Commission will publish the regulation’s final text and begin the process of setting the official publication date. Stakeholders in the AI sector should monitor the publication schedule and prepare to adjust compliance timelines accordingly.
The regulation’s passage marks a significant step in the EU’s effort to create a clear, simplified AI regulatory framework. While the core AI Act remains unchanged, the new timelines and clarifications provide additional flexibility for providers, regulators, and industry participants to align with the EU’s high‑risk AI obligations.
The regulation postpones the entry into force of several high‑risk AI obligations that were originally scheduled for 2 August 2026. Under the new timetable, stand‑alone high‑risk AI systems will become subject to the AI Act on 2 December 2027, while high‑risk AI embedded in products will be regulated from 2 August 2028.
Key provisions of the new regulation include:
A prohibition on AI systems that generate non‑consensual sexual or intimate content, including child sexual abuse material (CSAM). The ban covers AI that creates nude images of real people or edits photographs to remove clothing and reveal intimate parts. The prohibition takes effect in December 2026. A postponement of the deadline for national authorities to establish AI regulatory sandboxes. The new deadline is 2 August 2027, one year later than the original 2 August 2026 target. A reduction of the grace period for AI providers to implement transparency solutions for artificially generated content. The period is cut from six months to three months, with the new deadline set for 2 December 2026. Clarification of the European AI Office’s supervisory role over general‑purpose AI models that are developed and deployed by the same provider. The Office’s competence is limited where national authorities retain jurisdiction, such as law enforcement, border management, judicial authorities, and financial institutions. A mechanism to resolve overlaps between the AI Act and sector‑specific legislation. In cases where sectoral law contains similar AI‑specific requirements—examples include medical devices, toys, lifts, and watercraft—the AI Act’s high‑risk provisions can be limited through implementing acts. Exemption of products covered by the machinery regulation from direct applicability of the AI Act. The Commission may adopt secondary legislation to add health and safety requirements for high‑risk AI systems that fall under the machinery regulation. * A new obligation for the Commission to issue guidance that helps operators of high‑risk AI systems, which are also subject to sectoral harmonisation legislation, comply with the AI Act while minimising administrative burden.
The regulation will be published in the EU’s Official Journal shortly. It will enter into force on the third day after publication.
The move follows a series of EU initiatives aimed at simplifying digital regulation. In October 2024 the European Council urged all EU institutions and member states to accelerate work on simplification, citing reports by Enrico Letta and Mario Draghi. The Budapest declaration of 8 November 2024 called for a “simplification revolution” to reduce regulatory and reporting burdens, especially for small and medium‑sized enterprises. Since February 2025 the Commission has released ten Omnibus packages covering sustainability, investment, agriculture, digitalisation, defence, chemicals, automotive, food safety, and more.
The high‑risk AI delay aligns with the broader Omnibus VII agenda, which keeps the core architecture of the AI Act unchanged while providing additional time for providers and regulators to prepare. The postponement does not alter the AI Act’s core prohibitions or transparency obligations for general‑purpose AI.
The regulation’s ban on non‑consensual sexual content and CSAM reflects the EU’s ongoing effort to address harmful uses of generative AI. The AI Office’s expanded oversight role and the new guidance obligation aim to balance innovation with consumer protection.
In the coming weeks, the Commission will publish the regulation’s final text and begin the process of setting the official publication date. Stakeholders in the AI sector should monitor the publication schedule and prepare to adjust compliance timelines accordingly.
The regulation’s passage marks a significant step in the EU’s effort to create a clear, simplified AI regulatory framework. While the core AI Act remains unchanged, the new timelines and clarifications provide additional flexibility for providers, regulators, and industry participants to align with the EU’s high‑risk AI obligations.
