1. What's Happening
On June 29, 2026, the EU Council gave final adoption to the Digital Omnibus on AI, completing Parliament's June 16 endorsement. Per TechTimes (10 July 2026), high-risk AI system deadlines are now pushed back significantly: stand-alone systems (Annex III — employment, credit, biometrics, law enforcement) until December 2, 2027, and product-embedded systems (Annex I) until August 2, 2028.
Two things are not delayed. Article 50 transparency duties — AI-interaction disclosure, synthetic-content marking, deepfake labeling — become enforceable August 2, 2026 (existing systems get until December 2, 2026 for marking). And per the same reporting, the Commission's power to fine GPAI providers also activates August 2, 2026, applied retroactively to violations back to August 2025 — though other sources describe this power as prospective only from August 2026, and which reading is correct remains unsettled as of this writing.
2. The Principle: Deferral for Deployment, Not for Disclosure or Enforcement
The Omnibus buys time on the Act's most demanding piece — high-risk compliance — while holding the line on transparency and enforcement. Reading the deadline extensions as a broad reprieve risks missing that disclosure duties and the Commission's toolkit against foundation-model providers are unaffected, and active within weeks.
OUR FORECAST
We expect the first real GPAI enforcement action to also resolve whether “August 2, 2026” reaches back to 2025 conduct or only looks forward. Organizations that assumed one reading should not wait for that test to find out which was correct.
3. What We Recommend Now
- Separate your Annex I and Annex III system inventories now that their compliance deadlines diverge by eight months.
- Confirm Article 50 disclosure and content-marking compliance for any customer-facing AI system by August 2, 2026, regardless of your high-risk classification timeline.
Ne'eman Keynan applies the same transparency principle internally: every AI-assisted client draft carries a mandatory disclosure statement on the firm's AI system's role, reviewed and approved by a supervising attorney before it reaches you.
Source: TechTimes, 10 July 2026
This update is provided for general informational purposes and does not constitute legal advice. For guidance specific to your organization's use of AI systems, please contact Philippe Lipschutz, Adv..
This document combines the firm's advanced AI system with experienced legal expertise, under the close supervision and approval of senior attorney.
עדכונים ששייכים ל- LEGAL BRIEFS: