The European Parliament has reached its own preliminary position on EU AI Act amendments. IAPP confirmed the development with a direct headline: “MEPs reach preliminary political agreement on AI omnibus.” The agreement was reached on or around March 11, 2026.
This is Parliament-side movement. Prior coverage on this hub addressed the Council’s amendment package, the nudification ban addition, the three-track structure. This is a distinct institutional step: the Parliament’s committee representatives have now aligned on their amendment position. Both institutions must ultimately agree for amendments to become law.
The specific provisions require careful framing. According to IAPP reporting, the agreement reportedly extends compliance deadlines for high-risk AI systems, Annex III requirements to December 2, 2027, and Annex I requirements to August 2, 2028. The deal reportedly shortens the grace period for generative AI transparency requirements to three months, per IAPP. These dates and provisions could not be independently confirmed from the article body, which was partially accessible. Until the March 18 committee vote produces official documentation, treat these figures as reported, not final.
The generative AI timeline compression is the more immediate concern for many operators. A shortened grace period on transparency requirements means less runway for compliance, and it runs counter to the direction of the deadline extensions elsewhere in the package.
The March 18 committee vote will either advance or modify this preliminary agreement. If you’re managing an EU AI Act compliance roadmap, that vote warrants same-day attention. Check back here for updated coverage following the vote.