NeuralStackThe European Union’s AI Omnibus has entered into force, introducing targeted changes to the bloc’s AI rulebook and giving companies more time to prepare for several obligations. The package was proposed on 19 November 2025 as part of a broader digital omnibus initiative and is intended to simplify implementation without weakening protections for safety and fundamental rights.
One of the most consequential changes is a revised timetable for high-risk AI systems. Systems listed under Annex III will face the relevant rules from 2 December 2027, while high-risk AI embedded in physical products such as machinery, toys and lifts will be covered from 2 August 2028.
The regulation also expands support for companies developing AI in Europe. Measures previously reserved for small and medium-sized enterprises will extend to small mid-cap companies, allowing more businesses to benefit from proportionate compliance requirements.
Testing opportunities are being widened as well. The Omnibus expands access to regulatory sandboxes and introduces an EU-level sandbox where companies can develop and test AI systems under regulatory supervision.
Administrative requirements have been reduced in several areas. Simplified legal obligations for SMEs now extend to small and mid-cap companies, the corporate AI literacy requirement is streamlined, and registration requirements for exempted AI systems in the EU’s central database are simplified.
The changes retain and add safeguards in sensitive areas. The Omnibus prohibits AI systems that generate non-consensual sexually explicit or intimate content or child sexual abuse material, while allowing special categories of personal data to be processed for the purpose of detecting and correcting bias.
Why this matters
For AI developers and deployers, the package trades some immediate compliance pressure for a longer preparation window and clearer procedures. Smaller companies may find it easier to participate in the European market, while supervised sandboxes could help them test products before facing full regulatory obligations.
The changes also reshape enforcement. The AI Office receives extended oversight of certain systems, including AI built on general-purpose models and systems embedded in large online platforms and search engines. The Omnibus further clarifies how the AI Act interacts with other EU laws and simplifies procedures for conformity assessment bodies.
Companies now have additional time to prepare for the delayed high-risk rules, while regulators and member states take on a stronger role in supporting AI literacy and implementation. The result is a more gradual rollout that aims to balance innovation and competitiveness with continued protection of people’s rights and safety.