According to a statement released by the Council of the European Union on May 7, 2026, the Presidency of the Council and negotiators from the European Parliament have reached a provisional agreement on a proposal to simplify AI-related rules. The proposal forms part of the EU’s “Omnibus VII” package aimed at simplifying digital regulations, with one of its key focuses being the adjustment of the application timeline for certain provisions of the AI Act.

The most significant postponement under the provisional agreement concerns the rules applicable to high-risk AI systems. The European Commission had previously proposed delaying the application of the high-risk AI system rules by up to 16 months, so that the relevant obligations would only begin to apply once the necessary standards and implementation tools are in place. Since the provisions concerning high-risk AI systems under the AI Act were originally scheduled to apply from August 2, 2026, the EU co-legislators treated the proposal as a priority matter.

Under the provisional agreement, new fixed application dates will be established for the high-risk AI system rules. Specifically, the rules concerning stand-alone high-risk AI systems are proposed to be postponed until December 2, 2027, while the rules concerning high-risk AI systems embedded in products are proposed to be postponed until August 2, 2028.

In addition, the provisional agreement would extend the deadline for national competent authorities to establish AI regulatory sandboxes until August 2, 2027.

Apart from postponing certain provisions, the agreement also shortens the grace period for some obligations. The grace period for providers to implement transparency measures for artificially generated content would be reduced from six months to three months, with the new deadline set for December 2, 2026.