The Digital Omnibus on AI — the first major legislative revision to the EU AI Act since its adoption — has completed the legislative process. The Council and the European Parliament reached provisional political agreement on 7 May 2026; the Parliament formally endorsed the text on 16 June; and the Council gave its final green light on 29 June 2026. The act is published in the EU's Official Journal and enters into force on the third day after publication, in July 2026.
The Omnibus reshapes the compliance calendar rather than the substance of the obligations. Stand-alone high-risk systems under Annex III now apply from 2 December 2027, and high-risk AI embedded in regulated Annex I products from 2 August 2028 — extensions intended to give providers and deployers more time for the Act's technical documentation, risk-management and conformity-assessment work.
Beyond the deadlines, the package recalibrates several provisions. The Article 4 AI-literacy duty is softened from a mandatory obligation to an encouragement. The simplified documentation regime originally reserved for SMEs is extended to small mid-cap companies, widening the pool that qualifies for lighter paperwork. And the EU AI Office gains expanded competence to supervise AI systems integrated into very large online platforms and search engines.
Critically, the Omnibus did not move the 2 August 2026 date for Article 50 transparency obligations and GPAI enforcement powers — those remain live. Organisations should read the Omnibus as breathing room on the heaviest high-risk workstreams, not as a reprieve from the transparency and general-purpose-AI regime taking effect this summer.
Sources: Council of the EU, "Artificial Intelligence: Council and Parliament agree to simplify and streamline rules" (7 May 2026); Freshfields, "EU AI Act unpacked #34: The final Digital Omnibus on AI".