Become a partner
DE EN
NADOVO

Digital Omnibus: More Time, Not Less Obligation

The Digital Omnibus pushes central EU AI Act deadlines to December 2027 - but postponed isn't cancelled. What gets postponed, what actually gets tighter, and why the extra time is a preparation window, not a cushion to rest on.

News & Updates

Digital Omnibus: More Time, Not Less Obligation AI-generated

“The EU has watered down the AI Act. So we can just wait and see for now.”

Postponed isn’t cancelled. Anyone who waits now mistakes a deferral for a free pass.

In the Digital Omnibus package, the EU decided in May 2026 to postpone central deadlines of the EU AI Act and to ease individual obligations. That quickly becomes: it’s all half as bad. In fact the timeline shifts, but the direction stays. And some things even get tighter.

What gets postponed

The big chunk first. The comprehensive obligations for high-risk AI under Annex III now apply from 2 December 2027 instead of, as before, from August 2026. For high-risk AI embedded as a safety component in regulated products (Annex I), the date moves to 2 August 2028.

There’s more breathing room elsewhere too. The national AI regulatory sandboxes only need to be in place by 2 August 2027. And the AI literacy obligation under Article 4 is softened: instead of ensuring a certain level, companies will in future have to take measures to promote it.

What gets tighter, not looser

Here comes the part the wait-and-see camp overlooks. Not everything is being relaxed.

The grace period for watermarking AI-generated content (Article 50) is shortened from six to three months. The new deadline is 2 December 2026. Anyone who hoped for a long transition period gets less time, not more.

What isn’t postponed at all

Two things have applied unchanged since 2 February 2025: the bans on certain AI practices and the AI literacy obligation. The latter is softened, but not abolished, and continues to apply in its original wording until the final reform.

And the transparency obligations under Article 50 still take effect from 2 August 2026. Anyone who uses AI for communication and content therefore needs to deal with it not in 2027, but in a few weeks.

The penalties also remain untouched. Prohibited practices can incur up to 35 million euros or 7 percent of global annual turnover, other violations up to 15 million or 3 percent. The Omnibus lowers the effort, not the risk.

State of play

Important for context: the reform isn’t in force yet. As of 17 June 2026.

The political agreement of 7 May was followed on 16 June by the European Parliament’s approval. Still missing are the formal adoption by the Council and publication in the EU Official Journal. Both are expected in July 2026, in time before the old August deadline. Until publication, the original timeline formally applies.

For practice that means: plan with the new deadlines, but don’t rely on them until they’re in black and white in the Official Journal.

Why the postponement is no reason to sit back

The stated goal of the Digital Omnibus is to reduce administrative burden, by at least 25 percent by 2029, and for small companies even by 35 percent. Good news for mid-sized businesses.

But less effort is not an exemption from the obligation. The extra time is a preparation window, not a cushion to rest on. Anyone who wants a clean AI register, clear risk classes and documented measures by December 2027 doesn’t start in November 2027. The difference between “we have time” and “we use the time” decides whether the deadline ends up stress-free or hectic.

The first step stays the same anyway, no matter which date applies: knowing which AI is even in use. That’s exactly what the NADOVO platform makes visible. What the structured entry looks like is shown in our 4-week plan. And anyone who needs support with implementation will find it in our AI compliance consulting.

One question to close. When the new deadlines apply: does that start a break for you, or a preparation window? Where you stand today is shown by our quick check.


About the author

Jochen Stier is a co-founder of NADOVO with over 20 years of experience in process management and IT service management. He helps German SMEs implement the requirements of the EU AI Act systematically and pragmatically. His 5-phase NADOVO framework combines regulatory requirements with practical feasibility, without enterprise budgets or complex tools.

Further reading:


← Back to blog

© 2026 CONTORO SOLUTIONS OÜ. All rights reserved.
NADOVO - nadovo.ai