Council of the EU formally adopts Digital Omnibus on AI

The Council of the EU formally adopted the Digital Omnibus on AI, closing the legislative process that had begun with a provisional deal in May and cleared the European Parliament on 16 June. The package amends the EU AI Act to push back the compliance deadlines for high-risk systems, which had originally been due to apply from August 2026.

Under the revised timetable, standalone high-risk systems covered by Annex III — including those used in education, employment screening, credit scoring, biometric identification and critical infrastructure — now have until December 2027 to comply, while high-risk AI embedded within other regulated products has until August 2028. Not every deadline moved: the Act’s separate transparency obligations, including the requirement that providers disclose when a user is interacting with an AI system, still take effect in August 2026 as originally scheduled.

The Omnibus enters into force twenty days after publication in the EU’s Official Journal. Supporters, including the Commission, had argued the extension gave companies and regulators more realistic lead time to build the compliance infrastructure the Act required; critics saw it as the EU softening its flagship AI legislation under industry and diplomatic pressure less than two years after the Act’s core provisions began taking effect. The adoption left the AI Act’s basic structure intact while significantly loosening its near-term bite.