EU Council agrees position to simplify implementation of AI rules
EU member states have agreed a negotiating position on proposals to streamline certain artificial intelligence regulations and adjust the timeline for implementing parts of the AI Act.
The Council of the European Union has agreed its position on a proposal to simplify aspects of the EU’s AI regulatory framework. The initiative forms part of the ‘Omnibus VII’ package, which aims to streamline parts of EU digital legislation and support the implementation of the AI Act.
The proposal includes adjustments to the timeline for applying rules on high-risk AI systems. The European Commission suggested delaying their application by up to 16 months to ensure that technical standards and implementation tools are available before the rules take effect.
Under the Council’s position, the new deadlines would be 2 December 2027 for stand-alone high-risk AI systems and 2 August 2028 for high-risk AI systems embedded in products.
The Council also introduced additional provisions, including a prohibition on AI systems used to generate non-consensual intimate content or child sexual abuse material. Other amendments clarify the role of the EU AI Office and require companies to register AI systems in the EU database even when they claim their systems are exempt from high-risk classification.
The text also postpones the deadline for establishing national AI regulatory sandboxes to December 2027 and allows limited processing of sensitive personal data to detect and correct bias in AI systems.
With the Council’s mandate now approved, negotiations with the European Parliament will begin to finalise the legislative changes.
