EU AI Omnibus enters into force with simpler implementation rules

The European Union’s AI Omnibus entered into force on 27 July 2026, introducing targeted amendments to the AI Act that simplify implementation, extend compliance timelines and strengthen support for AI innovation while maintaining core safeguards for safety and fundamental rights.

The legislation forms part of the broader Digital Omnibus package proposed in November 2025. It reduces administrative burdens for businesses, particularly smaller companies, while expanding regulatory support for AI innovation across the EU.

Among the main changes, the AI Omnibus broadens access to regulatory sandboxes by creating an EU-wide sandbox and extending existing support measures from SMEs to small mid-cap companies.

It also delays key compliance deadlines, with obligations for Annex III high-risk AI systems applying from 2 December 2027 and requirements for AI integrated into regulated physical products, including machinery, toys and lifts, taking effect from 2 August 2028.

The legislation introduces additional simplifications, including streamlined AI literacy obligations, reduced registration requirements for exempt AI systems in the EU’s central database and clearer conformity assessment procedures.

At the same time, it strengthens safeguards by prohibiting AI systems used to generate non-consensual intimate content or child sexual abuse material and by permitting the processing of sensitive personal data where necessary to detect and mitigate algorithmic bias.

The AI Omnibus also expands the supervisory role of the EU AI Office over certain general-purpose AI systems and AI deployed within very large online platforms and search engines.

According to the European Commission, the reforms are intended to improve legal certainty, align the AI Act more closely with other EU legislation and encourage innovation without weakening its core objectives.

Why does it matter?

The AI Omnibus marks the EU’s first major adjustment to the AI Act since its adoption, signalling a shift from designing AI regulation to refining its practical implementation. By extending compliance timelines and simplifying administrative requirements, the EU is seeking to improve regulatory predictability while preserving protections for high-risk AI applications and fundamental rights.

The reforms also illustrate a broader evolution in AI governance. Rather than weakening regulation, policymakers are increasingly focused on making complex regulatory frameworks easier to implement, particularly for smaller businesses and public authorities.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!