Brussels: The European Union’s Digital Omnibus entered into force on 27 July 2026, three days after its publication in the Official Journal, reshaping how the bloc applies the Artificial Intelligence Act and postponing the toughest obligations for high-risk AI systems. The regulation streamlines rules that many businesses said were arriving faster than the technical standards meant to support them.
The Council gave the package its final green light on 29 June 2026, Parliament adopted it on 16 June, and the final act was signed on 8 July. The Commission first proposed the Digital Omnibus on 19 November 2025 as part of a wider drive to simplify the EU’s digital rulebook.
The headline change delays the AI Act’s high-risk requirements. Stand-alone systems listed in Annex III, covering areas such as employment, education and access to essential services, now face a compliance deadline of 2 December 2027 instead of 2 August 2026. AI embedded in regulated products under Annex I gains still more time, until 2 August 2028.
Henna Virkkunen, the Commission’s Executive Vice-President for Tech Sovereignty, Security and Democracy, framed the reform as a bid to keep Europe competitive without lowering standards.
“We need to make doing business in Europe easier without compromising our high standards of online fairness and safety. We want an innovation-friendly rulebook: both in the way we apply the rules, and in simplifying the laws where our objectives can be reached at lower costs and streamlined procedures,” Virkkunen said.
The Digital Omnibus does more than push deadlines. It hands the Commission’s AI Office broader powers to supervise general-purpose AI models and links that oversight to platforms already regulated under the Digital Services Act. The law also writes a new prohibition into the AI Act, banning systems designed to generate non-consensual sexual or intimate imagery and child sexual abuse material.
Alongside the AI provisions, the package trims reporting duties across neighbouring files, amending the EU’s civil aviation rules and its machinery regulation to remove overlapping paperwork. Officials argue the changes hand companies legal certainty before the hardest obligations bite.
Critics are less convinced. Digital rights groups, including a coalition that wrote to Virkkunen during the drafting, warned that reopening the AI Act barely a year after its adoption risks diluting protections and rewarding lobbying. They fear repeated deferrals could become a template for weakening other digital laws.
Industry groups counter that the original timetable was unworkable because harmonised standards, guidance and testing infrastructure were not ready. For them, the Digital Omnibus buys time to build compliance systems rather than a licence to ignore the rules.
The AI Act, agreed in 2024, remains the world’s first comprehensive law governing artificial intelligence. It sorts systems by risk, bans a handful of practices outright, and imposes transparency and safety duties on the rest. Its staggered timeline always meant different obligations would apply on different dates, and the new regulation resets several of those dates rather than the framework itself.
Readers can consult the Council’s final approval statement and the Commission’s AI framework pages for the full text and implementation timeline. National authorities must now update guidance so companies know which obligations apply, and when, as Europe recalibrates its flagship technology law.
