On 20 April 2026, the Dutch government published its proposal for the AI Act implementing legislation (the Uitvoeringswet verordening artificiële intelligentie, or “Uvai”) for public consultation. Because the AI Act (Regulation (EU) 2024/1689) is directly applicable, the implementing bill focuses primarily on establishing the national framework necessary for supervision and enforcement, with the stated objective of ‘pure and low-burden’ implementation, meaning no additional substantive rules unless strictly necessary, and minimal administrative burdens where Member State discretion exists. Notably, the Netherlands has not used the limited exceptions in the AI Act to introduce additional national requirements beyond what the Regulation itself mandates.
The proposed supervisory framework
The draft Uvai primarily focuses on the designation of competent authorities, largely assigning responsibility to regulators already charged with oversight of the relevant product or sector. The principal allocations are proposed as follows:
- Coordination and single point of contact: The Dutch Data Protection Authority (Autoriteit Persoonsgegevens, or “AP”) and the Dutch Authority for Digital Infrastructure (Rijksinspectie Digitale Infrastructuur, or “RDI”) will be jointly responsible for coordinating the national AI supervisory system, including alignment of supervisory approaches, information sharing, and knowledge-building across all designated authorities. In addition, the RDI acts as the Single Point of Contact under Article 70(2) of the AI Act, and must fulfill this role in close coordination with the AP.
- Prohibited systems, transparency risks and Annex III: The AP is set to take on a wide range of responsibilities, supervising prohibited AI practices, transparency obligations, and the majority of high-risk AI applications under Annex III of the AI Act, including biometrics, education, recruitment, access to essential services (excluding financial institutions), and law enforcement and migration.
- Financial sector: The Authority Financial Markets (Autoriteit Financiële Markten, or “AFM”) and the Dutch National Bank (De Nederlandse Bank, or “DNB”) are proposed to be designated as the competent supervisory authorities for AI systems used by financial institutions in areas such as creditworthiness assessment, as well as for prohibited AI practices and transparency obligations in the financial sector.
- Critical infrastructure: The RDI and the Human Environment and Transport Inspectorate (Inspectie Leefomgeving en Transport, or “ILT”) will be responsible for high-risk AI systems used within critical infrastructure, including critical digital infrastructure, energy, road, and drinking-water sectors.
- Product-related AI (Annex I): Existing product market surveillance authorities, including RDI, ILT, the Health and Youthcare Inspectorate (Inspectie Gezondheidszorg en Jeugd, or “IGJ”), the Netherlands Labour Authority (Nederlandse arbeidsinspectie, or “NLA”) and the Netherlands Food and Consumer Product Safety Authority (Nederlandse Voedsel- en Warenautoriteit, or “NVWA”), will retain their designated roles for AI systems embedded in regulated products.
- Administration of justice: Given the constitutional principle of judicial independence, oversight of AI systems used in courts will be allocated internally to the Procurator General at the Supreme Court and the Chair of the Administrative Jurisdiction Division of the Council of State (Afdeling bestuursrechtspraak van de Raad van State), with standard enforcement powers modified accordingly.
Regulatory sandbox
Under the AI Act, Member States are required to establish at least one AI regulatory sandbox by 2 August 2026. In the Netherlands, under the current draft Uvai, all eight designated market surveillance authorities will jointly operate a single, multi-sectoral sandbox, coordinated by the RDI and AP, with a digital portal for applications and general enquiries.
Key takeaways for organisations
For businesses operating or marketing AI systems in the Netherlands, the primary relevance of this proposal lies not in the introduction of new substantive obligations, but in clarifying the supervisory framework. Under the draft Uvai, supervision is expected to be distributed across sectoral regulators, meaning that organisations will, in most cases, be overseen by authorities with which they are already familiar.
The consultation period presents a valuable opportunity to provide input on the proposed supervisory architecture, coordination mechanisms, and enforcement powers before these are finalised in law. Stakeholders may submit their responses until 1 June 2026.
For further information, please contact:
Puck van den Bosch, Bird & Bird
puck.vandenbosch@twobirds.com
