The EU Regulation on Artificial Intelligence (AI) entered into force on 2 August 2024. The EU rules were followed by certain national rules on how the Danish authorities can enforce the new requirements in practice. We have previously written about that here and here.
These new AI rules are gradually entering into force, and the draft bill is just adding another procedural layer. More specifically, the rules give the relevant authorities the power to implement certain measures in practice in the event of non-compliance. In such cases, companies may face obligations to:
- Warn users about the risks of using the system
- Stop misleading marketing activities
- Recall the system from the users
Pursuant to the draft bill, companies may also face prohibitions on supplying, marketing, and more when an AI system triggers a risk. Certain AI systems must also be registered with the Danish Agency for Digital Government. That, namely, includes systems used for critical digital infrastructure.
The aim is for the draft bill to enter into force on 2 August 2026.
iuno's opinion
There are no new obligations triggered for companies with the draft bill. Instead, it confirms the framework for enforcing the existing and coming rules on AI. That just emphasizes the importance of companies internally identifying the responsibility for the relevant AI systems. Companies may also become liable as importers, distributors, or operators of such systems.
iuno recommends that companies carefully map the use of AI. Such mapping should include a clear categorization of the system’s risk category. It also entails an ability to define what data the system is based on and how it works in practice. That is especially important within the HR department, where AI systems may have far-reaching consequences.
[Draft bill on supplementary rules to the regulation on artificial intelligence of 18 February 2026]