Regulation (EU) 2026/1744

The AI Act digital omnibus: what changed in the timeline and the obligations

Regulation (EU) 2026/1744 of 8 July 2026, published in the Official Journal on 24 July and in force since 27 July 2026, amends the AI Act before its heaviest obligations kick in. It delays the high-risk regime, adds prohibitions and eases some requirements. Here is what moved, date by date.

By Younes Lahrach, data protection specialist and external DPO, director of Sila International Advisory. First published on , last reviewed against the texts on .

The dates that moved

AI Act timeline before and after Regulation (EU) 2026/1744
Annex III high-risk systems (Article 6(2))2 August 20262 December 2027Regulation (EU) 2026/1744, Article 113(c)(i) as amended
High-risk systems linked to Annex I products (Article 6(1))2 August 20272 August 2028Regulation (EU) 2026/1744, Article 113(c)(ii) as amended
New prohibitions in Article 5(1)(ba) and (bb)Not applicable2 December 2026Regulation (EU) 2026/1744, Article 113(a) as amended
Marking of synthetic content (Article 50(2)), systems placed on the market before 2 August 20262 August 20262 December 2026Regulation (EU) 2026/1744, Article 111(4)

Checked on EUR-Lex. Other milestones (original prohibitions from 2 February 2025, general-purpose AI models from 2 August 2025, Article 50 transparency from 2 August 2026) are unchanged.

What did not move

Article 5 prohibitions have applied since 2 February 2025. Obligations on general-purpose AI model providers have applied since 2 August 2025, and Article 111(3) still gives models placed on the market before that date until 2 August 2027. Article 50 transparency obligations still apply from 2 August 2026.

So the high-risk delay is not a general delay. A chatbot or content generator must already meet Article 50.

Substantive changes

AI literacy, Article 4

Providers and deployers must now take measures to foster AI literacy among their staff. The new text states that this does not oblige them to guarantee a specific level of AI literacy for any individual. The Commission and Member States support the effort, especially for SMEs.

Sensitive data and bias, Article 4a

The legal basis for processing special categories of data to detect and correct bias, previously in Article 10(5), is removed from that article and moved to a new Article 4a.

Machinery, Annex I

Point 1 of Section A of Annex I, on machinery, is deleted and moved to Section B.

Fines for small mid-caps, Article 99(6a)

The lower-of-two-amounts rule, previously reserved for SMEs, is extended to small mid-cap enterprises for fines under Article 99(4) and (5).

What it means for projects under way

The extra time for high-risk systems should go into qualifying systems correctly, not into pausing work. Harmonised standards and guidelines for high-risk systems shape how conformity is demonstrated; the later they arrive, the narrower the implementation window.

For software vendors, two deadlines are close: 2 December 2026 for the new prohibitions and for marking content generated by systems already on the market.

Frequently asked questions

What this regime triggers elsewhere

The reference page with the consolidated timeline from 2024 to 2028.

The AI Act: obligations and full timeline

The seven-step audit for SaaS products built on a large language model.

How to audit an LLM wrapper for EU market access

About the author

Younes Lahrach is the founder and director of Sila International Advisory. A GDPR legal consultant and external data protection officer since July 2025, he holds a Master's degree (M2) in European Union law, specialising in data protection, from the University of Lille, and an MBA in International Finance from Middlesex University. He is a legal consultant, not a lawyer (avocat).

His background →

Reference texts

The information on this page is general in nature and does not constitute legal advice on a specific situation. Sila is not a law firm and does not provide representation before the courts.