The legislative window is closed. On June 29, 2026, the EU Council gave its final green light to the Digital Omnibus on AI, the package of amendments to the EU AI Act that the European Parliament had approved on June 16. With Council adoption complete, the revised deadline structure is enacted law, not pending guidance. Organizations that have been waiting on legislative clarity before committing to compliance roadmaps no longer have that excuse.
Why it matters
Four compliance deadlines now govern how the EU AI Act applies to your systems, and they aren’t uniform. Article 50 transparency obligations, the requirement that providers inform users they’re interacting with an AI system, are not delayed and hit on August 2, 2026. That’s 32 days from today. If you deploy a consumer-facing AI system in the EU, this deadline is live and enforceable regardless of what else shifted in the Omnibus.
What did shift: stand-alone high-risk AI systems under Annex III, covering education, employment screening, credit scoring, biometric identification, and critical infrastructure, now face a compliance date of December 2, 2027, rather than the original August 2, 2026. That’s a 16-month extension. Don’t treat it as a reprieve from planning; treat it as a planning window that’s already open.
Who This Affects
Context
Morgan Lewis characterized the amendments plainly: businesses should treat these changes “primarily as an extension of time to complete their AI Act compliance efforts, rather than as a material relaxation of the underlying obligations.” The obligations themselves haven’t changed. The deadlines moved. That distinction matters for governance documentation, because the underlying risk classification framework and conformity assessment requirements remain intact.
One carve-out worth flagging: Article 50(2) watermarking obligations for synthetic audio, image, video, and text systems that were already on the market before August 2, 2026 get until December 2, 2026 to comply. New deployments after August 2 don’t get that grace period. According to Morgan Lewis’s analysis, high-risk AI embedded as a safety component in regulated products faces a deadline of August 2, 2028, though that specific date is attributed to their analysis of the adopted regulation and should be confirmed against the final Official Journal text when published.
What to watch
Publication in the Official Journal is the final step before formal entry into force, Morgan Lewis noted this process was expected to complete by August 2, 2026. Watch for that publication date, because it starts the clock on any remaining transition provisions. Also watch the EU AI Office’s enforcement posture on Article 50 violations after August 2: the first enforcement actions under the transparency rules will set the interpretive tone for operator liability.
TJS synthesis
The catch is that the 16-month extension for stand-alone high-risk systems creates a false sense of runway. Annex III systems require risk assessments, technical documentation, conformity assessments, and, in many cases, notified body involvement. None of that happens in the final quarter before a deadline. Organizations that bank the extension and defer planning until mid-2027 will find themselves in the same position they were in before the Omnibus: scrambling. The real question is whether your Article 50 disclosure posture is ready for August 2, because that deadline didn’t move at all.