The digital omnibus on AI has been a hot topic of debate over the past year, especially since the European Commission released its proposal on November 19, 2025. Businesses utilizing artificial intelligence in the EU have been navigating the implications of the AI Act, deciphering which obligations remain, which have been postponed, and which have been altered.
For organizations utilizing computer vision or agentic AI systems in regulated sectors like manufacturing, healthcare, or transportation, the digital omnibus proposal holds significant relevance. It impacts compliance timelines, documentation requirements, and the labeling of AI-generated content moving forward. Here’s a breakdown of what has changed, when these changes occurred, and the practical implications.
TL;DR: The digital omnibus on AI maintains the EU AI Act’s goals but adjusts the timeline for enforcement, allowing regulators, notified bodies, and businesses more time to align with technical standards.
The digital omnibus on AI is a targeted legislative package amending Regulation (EU) 2024/1689, also known as the EU AI Act, alongside regulations on civil aviation and machinery. Published on November 19, 2025, the proposal aims to address implementation challenges and reduce the regulatory burden identified after the AI Act came into effect. It is part of a broader digital omnibus package covering data, cybersecurity, and privacy regulations.
The primary rationale behind the digital omnibus is to synchronize compliance deadlines with the availability of technical standards, conformity assessment bodies, and national competent authorities. By deferring high-risk AI obligations to December 2, 2027, and August 2, 2028, for embedded AI systems, the omnibus ensures that regulatory frameworks catch up with operational realities.
The path from the proposal to the entry into force of the digital omnibus took approximately eight months, following the standard EU legislative procedure. The swift turnaround reflects industry and member states’ pressure to address implementation uncertainties before the original August 2026 deadline.
Key changes to high-risk AI systems include extending the application timeline to align with the availability of harmonized standards. Providers and deployers now have more time to prepare for the obligations under the AI Act, such as risk management, documentation, and human oversight.
Transparency obligations and AI-generated content requirements remain on schedule, except for the machine-readable marking mandate, which has been delayed. New prohibited practices have been introduced, such as banning AI systems from generating abusive material without consent.
AI literacy obligations have been adjusted, with the responsibility for training staff remaining with businesses. The creation of a new regulatory category, small mid-cap companies (SMCs), extends compliance relief measures to companies that exceed SME thresholds but are not large corporations.
The digital omnibus also strengthens AI governance by expanding the AI Office’s oversight of platforms regulated under the DSA. This expansion clarifies jurisdiction over AI systems built on general-purpose models and those embedded in large online platforms.
For teams working on computer vision platforms and agentic AI systems, the digital omnibus alters the compliance timeline but maintains the core expectations. It is crucial to continue preparing for GDPR-compliant data practices, document model provenance, and uphold governance standards to avoid a last-minute rush when enforcement begins.
