AI Act Delays: Why Trustworthy AI Can’t Wait

The Digital Omnibus on AI will likely delay key milestones of the EU AI Act.
However, deadlines may move but responsibility doesn’t—build trustworthy AI now.

High-risk AI systems may now fall under the requirements from December 2027, and AI in regulated products from August 2028.
This creates additional time. It does not reduce responsibility.

What has not changed

The core structure of the AI Act remains intact.
Organisations still need to understand where their AI systems fall within the risk classification.
They still need to ensure transparency, documentation, and appropriate governance.
For high-risk AI systems, the requirements remain substantial. This includes data governance, human oversight, robustness, and technical documentation.

What this means in practice

The timeline shift is not a reason to pause.
It is an opportunity to build capability.
Organisations that use this time to conduct a gap analysis, strengthen internal understanding, and establish structured compliance approaches will be in a significantly stronger position.

The real question

AI compliance is not primarily a question of timing.
It is a question of how seriously organisations take their responsibility to design and deploy trustworthy systems.
Those who begin now will not only be compliant. They will be prepared.