The EU's AI Omnibus Enters Into Force, Easing Some AI Act Deadlines
The AI Omnibus took effect on 27 July 2026, simplifying parts of the EU's AI rulebook while pushing back deadlines for the highest-risk AI systems.
On 27 July 2026, the AI Omnibus entered into force across the EU, the Commission says, bringing extended timelines to what it calls the administrative simplification of the bloc's AI rules. The measure was first proposed as part of the wider digital omnibus package on 19 November 2025. The Commission describes it as a targeted simplification of the AI rulebook that still preserves strong safeguards for people's safety and fundamental rights.
Easier compliance and more room to test
The Commission says the Omnibus supports innovation and competitiveness in several ways: easing compliance for smaller businesses, extending timelines, expanding testing and experimentation opportunities, and giving companies developing and deploying AI in Europe greater legal clarity. Concretely, some measures that used to apply only to SMEs are now extended to small mid-cap companies, described as SMCs, so more businesses can benefit from a proportionate approach. The Omnibus also widens access to regulatory sandboxes and adds a new EU-level sandbox, so companies get more space to build and trial AI tools under regulatory oversight.
New deadlines for high-risk systems
The Omnibus pushes back two sets of deadlines for high-risk AI. Rules for high-risk AI systems listed in Annex III start applying on 2 December 2027. Rules for high-risk AI embedded in physical products covered by Annex I, such as machinery, toys, and lifts, start applying later still, on 2 August 2028.
Other simplifications
Beyond deadlines, the Omnibus simplifies several other obligations. It eases the previous AI literacy requirement for companies, while giving the Commission and Member States a stronger role in promoting AI literacy generally. It also simplifies the obligation to register exempted AI systems in the EU's central database.
Safeguards that stay in place
The Commission is clear that simplification does not mean fewer protections. The Omnibus bans AI systems that generate non-consensual sexually explicit or intimate content, or child sexual abuse material, sometimes called nudification apps. It allows the processing of special categories of personal data specifically to detect and correct bias in AI systems. It also extends the AI Office's enforcement powers, giving it oversight of certain AI systems, including those built on general-purpose models and those embedded in large online platforms and search engines. Finally, it clarifies how the AI Act interacts with other EU laws and simplifies procedures for conformity assessment bodies.
Key facts
- The AI Omnibus entered into force on 27 July 2026, having been proposed as part of the digital omnibus package on 19 November 2025.
- Rules for high-risk AI systems in Annex III apply starting 2 December 2027.
- Rules for high-risk AI embedded in physical products under Annex I apply starting 2 August 2028.
- The Omnibus extends simplified obligations previously reserved for SMEs to small mid-cap companies, and adds an EU-level regulatory sandbox.
- It bans nudification apps, allows bias-detection data processing, and extends the AI Office's enforcement powers.
Check your understanding
When do the rules for high-risk AI systems in Annex III start applying, and when do the Annex I rules for high-risk AI embedded in physical products start? What does the AI Omnibus do to protect people even as it simplifies compliance for smaller businesses?