On 7 May 2026 the European Parliament and the Council of the EU reached a provisional political agreement on the Digital Omnibus on AI. The package amends Regulation (EU) 2024/1689, the EU AI Act. The agreement is a provisional trilogue deal. Both co-legislators must still formally adopt the text. Formal adoption is targeted before 2 August 2026, the previous trigger date for the high-risk system rules.
The Omnibus rewrites the application timeline in Article 113 of the AI Act. Stand-alone high-risk AI systems listed in Annex III now apply from 2 December 2027. High-risk AI embedded in products covered by Annex I applies from 2 August 2028. The grace period for transparency and labelling of synthetic content under Article 50 shortens from six months to three months, with a new application date of 2 December 2026. Member State sandboxes under Article 57 are postponed to 2 August 2027. A new prohibition is added to Article 5, banning AI systems that generate non consensual intimate or sexual content and child sexual abuse material, with compliance required by 2 December 2026.
Providers of high-risk AI systems gain extra time to prepare conformity assessments, technical documentation, and post-market monitoring. Deployers of high-risk systems in HR, credit, insurance, biometrics, and critical infrastructure gain the same window. Foundation model providers and general-purpose AI model providers face revised Article 99 supervision rules. The AI Office gains primary jurisdiction where the high-risk system uses a general-purpose model built by the same provider, with carved exceptions for law enforcement, border management, judicial authorities, and financial supervision. Providers that generate or distribute synthetic media must meet labelling and watermarking duties by 2 December 2026.
The new Article 5 prohibition on AI generated CSAM and non-consensual intimate content has no transitional carveout for existing systems. Foundation model rules in Chapter V keep their 2 August 2026 application date with no Omnibus delay. The provisional text still needs European Parliament plenary adoption and Council approval. Substantive language can shift before publication in the Official Journal. The Commission may issue further implementing acts on technical standards for watermarking and bias testing.
Licentium advises providers, deployers, and importers of AI systems on EU AI Act readiness, and works with a partner network for cross-border filings. Work we undertake includes Article 6 high-risk classification, conformity assessment scoping, Annex III risk reviews, GPAI documentation, Article 50 transparency labelling, AI Office notifications, and contract uplift for AI vendors and resellers. Contact us to scope a review.
Source: Council of the European Union, Press release 299/26, Artificial Intelligence: Council and Parliament agree to simplify and streamline rules, 7 May 2026, https://www.consilium.europa.eu/en/press/press-releases/2026/05/07/artificial-intelligence-council-and-parliament-agree-to-simplify-and-streamline-rules/
The information provided is not legal, tax, investment, or accounting advice and should not be used as such. It is for discussion purposes only. Seek guidance from your own legal counsel and advisors on any matters. The views presented are those of the author and not any other individual or organization. Some parts of the text may be automatically generated. The author of this material makes no guarantees or warranties about the accuracy or completeness of the information.