Key Developments

The European Council has agreed to strengthen the AI Act with new provisions specifically prohibiting AI systems from generating non-consensual sexual and intimate content or child sexual abuse material. This update comes as part of broader efforts to streamline EU artificial intelligence rules, with the Council working to simplify implementation ahead of the August 2026 deadline when most remaining provisions become active.

Concurrently, the European Commission announced the €75 million EURO-3C Project aimed at building federated Telco-Edge-Cloud infrastructure to enhance European digital sovereignty. The Second Draft Code of Practice on marking and labelling of AI-generated content was also published on March 3, 2026, providing clearer guidelines for compliance.

Industry Context

These developments position the EU as the global leader in AI governance, setting standards that will likely influence regulation worldwide. The specific focus on non-consensual content addresses growing concerns about deepfake technology and AI-generated abuse material, issues that have become increasingly urgent as generative AI capabilities advance.

The digital sovereignty initiative reflects Europe’s strategic goal of reducing dependence on non-EU technology providers while maintaining competitive AI capabilities. This approach contrasts sharply with the more laissez-faire regulatory environments in other major markets.

Practical Implications

For Irish and European AI developers, these changes mean:

  • Compliance Requirements: Companies must implement robust content filtering and generation safeguards before August 2026
  • Labelling Obligations: AI-generated content must be clearly marked according to the new Code of Practice
  • Infrastructure Opportunities: The EURO-3C Project opens potential partnerships for cloud and edge computing providers
  • Market Positioning: EU-compliant AI systems may gain competitive advantage in regulated markets globally

The streamlined rules should reduce compliance complexity while maintaining strong protections, potentially benefiting smaller European AI companies that previously struggled with regulatory uncertainty.

Open Questions

Key uncertainties remain around implementation specifics: How will the content prohibition enforcement work in practice across different AI model types? What technical standards will define adequate labelling for AI-generated content? How will the digital sovereignty infrastructure integrate with existing cloud providers, and what role will Irish data centres play in this federated approach?

The timeline pressure also raises questions about industry readiness, particularly for smaller companies that may need significant technical adjustments to meet the August deadline.

Source: EU Council