The AI Act Timeline in Motion – What Does the Digital Omnibus Mean in Practice?
The timetable for the European Union’s AI regulation is potentially changing with the Digital Omnibus package. Although the changes are not yet legally in force, they may bring additional transition time and reduce obligations – particularly regarding the requirements applying to high-risk AI systems.
Text by Teemu Moilanen & Elisa Laatikainen, 12.5.2026 | Photo by Adobe Stock Photos
The European Union’s AI Act (Artificial Intelligence Act) is progressing in phases. The next significant phase of application is expected to begin on 2 August 2026, when a broad range of obligations concerning high-risk AI systems become applicable – including requirements relating to risk management, documentation, quality management, and conformity assessment.
In the health technology sector, the timeline has been somewhat different. AI-enabled medical devices and in vitro diagnostic devices frequently fall within the high-risk category linked to product safety legislation, whose original application date was set for August 2026.
However, the discussion around implementation has shifted rapidly during the spring of 2026. The backdrop is the so-called Digital Omnibus package, which aims to ease and simplify the obligations applying specifically to high-risk AI systems.
This change is connected to a broader EU-level effort to strengthen European competitiveness, reduce regulatory overlap, and accelerate the adoption of AI within businesses.
The Current Situation
Negotiators from the European Parliament and the Council have reached a preliminary political agreement on the Digital Omnibus package. This strongly suggests that the original AI Act timeline will change – however, the change is not yet legally binding or in force.
Entry into force requires formal approval from the European Parliament and the Council, as well as publication in the Official Journal of the EU. A pivotal moment came in May 2026, when negotiators reached a trilogue agreement on a package to simplify the AI Act.
The discussions have highlighted in particular:
- Postponement of obligations for high-risk AI systems
- Reduction of overlaps between sector-specific regulation and the AI Act
- Simplification of certain documentation and conformity requirements
- Clarification of rules on watermarking and deepfakes
Regarding high-risk systems, the current August 2026 deadline appears likely to shift by over a year, with entry into force pushed back to 2 December 2027. For systems linked to product safety legislation – such as many AI-enabled medical devices and in vitro diagnostic devices falling under the MDR and IVDR frameworks – the application date would be moved to 2 August 2028.
In addition, the agreement proposes a narrowing of the so-called safety component concept. Under the proposed change, an AI function would not automatically be subject to high-risk obligations merely because it assists the user or optimises a product’s performance, provided that its failure or malfunction does not give rise to a health or safety risk.
At this stage, however, it must be emphasised that this is not yet adopted legislation.
The Situation May Feel Contradictory
From a business perspective, the situation may feel contradictory. The market expects a lighter regulatory burden and more time, yet it would equally be unwise to halt preparations, as the final content and timeline have not yet been confirmed.
In practice, the most sensible approach at present is to continue preparations aimed at AI governance and compliance, and to prioritise the use cases of greatest commercial importance. At the same time, it is worth actively monitoring the progress of the trilogue negotiations and refraining from disproportionate investments until the final legislative text has been confirmed.
Viewed more broadly, these developments speak to a shift in the priorities of EU AI policy. The debate increasingly emphasises competitiveness, investment attractiveness, and technological sovereignty, as well as the ability of European businesses to scale AI solutions in global competition.
The events of spring 2026 demonstrate that the AI Act is no longer seen merely as a regulatory undertaking, but is becoming a central element of Europe’s industrial and competitiveness strategy.
Summary of the situation
A summary of the AI Act and the situation for businesses in May 2026.
Is the original AI Act timeline for high-risk obligations still legally in force?
– Yes.
Does it currently appear likely to change?
– Probably yes.
Is the EU’s political direction currently clear? What about the views at national level?
– Yes, towards postponement and simplification.
Should businesses halt their compliance preparations?
– No.
Should businesses wait and anticipate that the timelines will change and that more time will be available?
– Yes, with reasonable confidence.
When can we expect clarity?
– The EU intends to resolve the matter before 2 August 2026.
