Posted by KEY CHANGES
Eléonore SCARAMOZZINO, Avocate Constellation Avocats
The European Parliamentary Research Service provided a state of the latest evolution of the AI digital omnibus in 2026 and what to expect from it. The latest discussions introduce targeted but impactful adjustments to the implementation of the Artificial Intelligence Act, focusing on feasibility, timing and administrative burden rather than changing its core risk-based structure. The initiative responds to concrete implementation challenges, particularly delays in harmonised standards, designation of authorities and availability of compliance tools.
The introduction of a new prohibited practices to prevent the development and use of Generative AI systems permitting the nudification of identifiable natural persons (and not only focusing on Deepfakes).
A major adjustment concerns the timeline for high-risk AI obligations. Instead of fixed application dates, the Omnibus introduces a conditional timeline, linking compliance obligations to the availability of technical standards and guidance. In practice, this pushes the effective application of high-risk requirements to late 2027 or even 2028, with a long-stop date ensuring eventual enforcement even if standards lag behind.
A key change relates to simplification of compliance obligations. The proposal reduces administrative burdens through streamlined documentation, clearer conformity assessment procedures and extended proportionality measures.
The Omnibus also strengthens testing and innovation mechanisms, particularly through the expansion of regulatory sandboxes and the introduction of more flexible real-world testing frameworks.
Another important adjustment concerns governance and centralisation, with a reinforced role for EU-level coordination (notably the AI Office) and efforts to clarify interactions between the AI Act and other EU legislation.
The latest adjustments reflect a shift from strict ex ante compliance toward more adaptive and implementation-driven regulation. While the core principles of the AI Act remain intact, the Omnibus introduces greater flexibility in timelines, testing and compliance pathways, seeking to ensure that the framework remains operational, proportionate and aligned with technological and market realities.
Parliament assigned the file to the IMCO and LIBE committees, appointing Arba Kokalari (EPP, Sweden) and Michael McNamara (Renew Europe, Ireland) as rapporteurs. The IMCO and LIBE committees adopted a joint report on the file on 18 March 2026 with 101 votes in favour, 9 against and 8 abstentions.
The report aligns with the Council’s proposal for fixed deadlines for the delayed application of rules governing high-risk AI systems. Additionally, it proposes a targeted ban on AI systems that generate sexual and intimate content without consent and reformulates the conditions for processing special categories of personal data for bias detection and correction. The report also reinstates a simplified registration requirement for non-high-risk AI systems, maintains the extension of the simplified documentation process for medium-sized companies with minor modifications on fines, and lightens the regulatory burden for AI systems embedded in products already subject to sector-specific rules. Furthermore, it modifies the transitional period for compliance with obligations to mark AI-generated content to three months.
Finally, it envisages that providers and deployers of AI systems should support the improvement of AI literacy among their staff and ensure that the Commission and Member States encourage and support AI literacy in society.
Text adopted by EP on 26.03.2026
REPORT on the proposal for a regulation of the European Parliament and of the Council amending Regulations (EU) 2024/1689 and (EU) 2018/1139 as regards the simplification of the implementation of harmonised rules on artificial intelligence (Digital Omnibus on AI)
P10_TA(2026)0098 Simplification of the implementation of harmonised rules on artificial intelligence (Digital Omnibus on AI) Amendments adopted by the European Parliament on 26 March 2026 on the proposal for a regulation of the European Parliament and of the Council amending Regulations (EU) 2024/1689 and (EU) 2018/1139 as regards the simplification of the implementation of harmonised rules on artificial intelligence (Digital Omnibus on AI) (COM(2025)0836 – C10-0304/2025 – 2025/0359(COD))1 (Ordinary legislative procedure: first reading