The situation
On 29 June 2026, the EU Council gave final approval to a targeted regulation under the Omnibus VII package that changed parts of the AI Act’s implementation timetable and scope. The adopted law moved the application date for stand-alone high-risk AI systems to 2 December 2027 and for high-risk AI systems embedded in products to 2 August 2028. It also pushed back the deadline for member states’ competent authorities to establish AI regulatory sandboxes to 2 August 2027. The Council said the act will enter into force on the third day after publication in the Official Journal, but the publication date had not been established in the available material as of 1 July 2026.
The same regulation reduced the grace period for providers to implement transparency solutions for artificially generated content from six months to three months, setting a 2 December 2026 deadline. It also added prohibited AI practices covering non-consensual sexual and intimate content and AI-generated child sexual abuse material. The Council said bans on systems that generate nude images of real people or digitally remove clothing from existing photos will apply from December 2026. The text also clarified supervision competences between the AI Office and national authorities in some general-purpose AI cases and created a mechanism to limit overlap with sectoral rules in areas including medical devices, toys, lifts, watercraft and machinery.
Why it matters
This changed binding compliance and enforcement timelines across the EU single market rather than rewriting the AI Act wholesale. The immediate effect is to give high-risk AI providers, manufacturers of AI-enabled products, and national regulators more time before the main high-risk obligations apply, while accelerating the deadline for some transparency measures on AI-generated content. That shifts compliance spending, enforcement sequencing and product-readiness planning across affected sectors.
The added prohibitions and clarified supervisory boundaries also sharpened the near-term regulatory perimeter. Providers exposed to synthetic sexual content risks now face a clearer path to earlier restrictions from December 2026, while companies operating across regulated product sectors may see less duplication where EU sectoral legislation already contains similar AI-specific requirements. The unresolved Official Journal publication date still matters operationally because it determines when the adopted regulation formally enters into force.
What changed
A May 2026 political agreement had signaled the timeline relief and rule changes, but the Council’s 29 June 2026 final approval converted them into enacted EU law pending Official Journal publication.
The assessment
Current judgement
Improving for regulated firms' near-term compliance timelines, but still operationally significant because the measure accelerates some transparency obligations and adds new prohibited practices from December 2026.
Key assumption
The assessment assumes no newer Official Journal publication or implementing decision has superseded the Council’s stated next steps as of 1 July 2026.
What could challenge this assessment
The exact start of the new legal timetable still depends on Official Journal publication, which was not established in the available source material.
Impacts
Markets
The final approval eased near-term compliance pressure for high-risk AI providers and manufacturers by pushing main obligations into 2027 and 2028, while bringing some transparency compliance forward to 2 December 2026.
Economic Pressure
The regulation reduced immediate regulatory pressure on firms exposed to the AI Act’s high-risk regime, but tightened constraints for providers handling AI-generated content through earlier transparency and new prohibited-content rules.
Political Stability
The Council’s final approval converted an earlier political agreement into adopted EU law, giving member states and regulators a clearer implementation sequence even though entry into force still depends on Official Journal publication.
What to watch
- When will the regulation be published in the Official Journal, triggering entry into force on the third day after publication?
- How quickly will member states and EU authorities translate the revised deadlines, sandbox extension and supervisory clarifications into concrete enforcement guidance for high-risk AI and sector-regulated products?
Evidence and uncertainty
Low confidence. Low confidence. 1 retained references from 1 reporting origins (Consilium). Claim-level independent corroboration has not been established. Confidence concerns the evidence for the reported development, not the probability of an outlook.
Geographic context
- 01Brussels, Belgium (city)
- 02European Union (region)
- 03Luxembourg, Luxembourg (city)