With the collaboration of Alice Gogni.
On 11 November, 2025, the European Commission published the Digital Omnibus Package, consisting of three proposals for EU regulation and communication on the Data Union Strategy, with the aim of simplifying and improving the effectiveness of the European digital strategy.
Among other things, the EU Commission adopted a proposal for a regulation amending the Artificial Intelligence Act[1] (the “AI Act”), involving several aspects (“Digital Omnibus on AI”).
On 10 February, the European Parliament published its proposed amendments to the text, and now it is up to the European Council to take an official position.
The Digital Omnibus on AI proposes targeted simplification measures to ensure the AI Act is implemented in a timely, smooth and proportionate manner, as implementation challenges have arisen regarding the effective application of key provisions of the AI Act. The most important provisions concern:
- High-risk systems: the EU Commission proposes subordinating the entry into force of the rules on high-risk AI systems to a Commission decision confirming the availability of measures to comply with the requirements set out in Chapter III of the AI Act (e.g. harmonized standards, common specifications and Commission guidelines), while granting a transitional period between the decision and the start date of the obligation. While the EU Parliament agreed to postpone the deadline for the applicability of the rules on high-risk AI systems (e. 2 August 2027 for high-risk systems listed under Annex III of the AI Act, and 2 August 2028 for others), it proposed that the applicability of such rules should not be linked to a Commission decision.
Additionally, AI systems falling within the scope of Article 6(3) of the AI Act (i.e. systems listed in Annex III as potentially high-risk but not classified as such due to not posing a significant risk to health, safety, or fundamental rights) should not be required to register in the EU database under Article 49(2) of the AI Act. Imposing registration requirements on these systems would create an excessive compliance burden.
- Extension of regulatory simplifications to small-to-medium-sized companies (SMCs): similarly to small-to-medium-sized enterprises (SMEs), SMCs, including start-ups, may provide the elements of the technical documentation specified in Annex IV in a simplified manner.
- AI literacy: rather than imposing a specific obligation on providers and deployers of AI systems, the Commission and the Member States should foster AI literacy. However, the EU Parliament has already expressed its opposition, proposing to maintain the obligation for providers and deployers, albeit proportionate to the activity carried out.
- Post-market monitoring: the EU Commission has proposed removing the obligation to adopt a harmonized post-market monitoring plan.
- AI Office supervision powers: the EU Commission proposed that the AI Office be identified as the only competent authority responsible for supervising AI systems based on general-purpose AI models or embedded in very large online platforms and search engines.
- Processing special categories of personal data: the Digital Omnibus of AI sets out the conditions for lawfully processing special categories of personal data for the purpose of detecting, preventing, and mitigating biases in high-risk AI systems. In line with the joint opinion adopted by the European Data Protection Board and the European Data Protection Supervisor[2], the EU Parliament proposed stricter conditions for processing.
Other proposed amendments concern the scope of real-world testing for high-risk AI systems and the introduction of an EU-level AI regulatory sandbox, which would be managed by the AI Office.
Finally, targeted amendments should be introduced to clarify the interaction between the AI Act and other EU legislation and adjust the Act’s procedures to improve its overall implementation and functioning.
The Digital Omnibus on AI is expected to be adopted by the end of 2026, and its provisions will be gradually applicable from 2027 and 2028.
[1] Regulation EU 1689/2024
[2] EDPB-EPDS, Joint opinion 1/2026 on the Proposal for a Regulation as regards the simplification of the implementation of harmonised rules on artificial intelligence
