The Council of the European Union has reached an agreement on a legislative proposal to streamline rules within the Artificial Intelligence Act, marking a significant step in the EU’s broader “Omnibus VII” simplification agenda. Representing a push for digital sovereignty and reduced administrative burdens, the proposal adjusts the timeline for applying rules to high-risk AI systems. Under the new Council mandate, these rules will now take effect on December 2, 2027, for stand-alone systems and August 2, 2028, for AI embedded in products, ensuring that necessary standards and tools are fully available before enforcement begins.
A key addition to the mandate is a strict prohibition on AI practices involving the generation of non-consensual sexual content or child sexual abuse material. The Council also seeks to extend regulatory exemptions originally meant for SMEs to small mid-caps and reinstates obligations for providers to register AI systems in the EU database if they claim an exemption from high-risk classification. To manage governance, the mandate clarifies the powers of the AI Office regarding general-purpose AI models while preserving the authority of national bodies in sensitive sectors like law enforcement and finance.
Furthermore, the Council has postponed the deadline for establishing national AI regulatory sandboxes until late 2027 and introduced a requirement for the Commission to provide specific guidance to help businesses navigate high-risk requirements with minimal burden. This agreement, which follows calls for a “simplification revolution” by EU leaders, now moves to negotiations with the European Parliament to finalize the digital legislative framework and bolster European competitiveness.
