Key Developments

The Council of the European Union yesterday (March 13, 2026) agreed its position on streamlining AI Act rules as part of the “Omnibus VII” legislative package. The most significant addition is a new prohibition on AI systems that generate non-consensual sexual and intimate content or child sexual abuse material - directly addressing a regulatory gap the European Commission acknowledged.

The Council also established fixed timelines for high-risk AI system compliance: December 2, 2027 for stand-alone systems and August 2, 2028 for embedded systems. Additionally, providers must now register exempted high-risk systems in the EU database, while AI regulatory sandbox establishment is delayed until December 2027.

Industry Context

This development comes amid growing concerns about harmful AI applications, particularly deepfake technology creating non-consensual intimate images. Irish MEP Michael McNamara has been vocal about implementation urgency, warning that excessive delays could “deprive the regulation of the momentum it needs to work.”

The timing is critical as the European Commission recently missed a February 2 deadline to provide Article 6 guidance on high-risk system obligations, creating uncertainty for AI operators across the EU.

Practical Implications

For AI system providers, the new timelines provide clarity but require immediate preparation. Stand-alone high-risk AI systems have less than two years for compliance, while those developing embedded systems have slightly more breathing room until August 2028.

The mandatory EU database registration for exempted high-risk systems means additional administrative overhead for companies claiming exemptions. However, the Council emphasizes this will “bring greater legal certainty” and “ensure more harmonised implementation across member states.”

Irish AI companies and those serving EU markets must now plan for these specific deadlines while preparing for the new content generation prohibitions.

Open Questions

Key uncertainties remain around the delayed Article 6 guidance and how the new content prohibitions will be technically defined and enforced. The ongoing trilogue negotiations between the Commission, Parliament, and national governments will determine final implementation details.

With Irish politicians calling for urgent action to “set aside politics” and ban harmful AI applications, the speed of final agreement remains crucial for regulatory momentum.

Source: Council of the European Union