The EU AI Act is the world’s first AI law. It entered into force in August 2024, though only some of the provisions are in effect to-date. Under the risk-based approach to AI regulation provided for under the legislation, some types and uses of AI are prohibited altogether. Some of the strictest regulatory requirements remaining concern ‘high-risk’ AI systems and ‘general purpose AI’ models ‘with systemic risk’.
As an EU regulation, as opposed directive, the EU AI Act has direct effect in each EU member state. However, additional national legislation is needed to address certain matters under the regulation, including responsibilities for enforcement – and associated powers and procedures. On Monday, the Dutch government opened a consultation on draft legislation for this purpose in the Netherlands – more than eight months after the 2 August 2025 deadline for implementing such a law.
Amsterdam-based Nienke Kingma of Pinsent Masons, who specialises in technology law, said: “This draft bill outlines, amongst other things, how AI compliance will be supervised nationally. It follows a hybrid model involving 10 market regulators: regulators will oversee AI compliance within their domains.”
“There will also be new supervision tasks for the Dutch data protection authority and the State Inspectorate of Digital Infrastructure. According to the draft bill, those authorities will be appointed as the coordinating supervisory authorities for AI in the Netherlands. In carrying out this role, they will work together with sectoral regulators such as the Health and Youth Care Inspectorate, the Netherlands Food and Consumer Product Safety Authority, and the Dutch Labour Inspectorate, among others,” she said.
Kingma said companies, citizens and other stakeholders now have an opportunity to submit comments on the draft legislation. The consultation closes on 1 June 2026. Government officials will then review the responses and could revise the draft before it is officially submitted to the Dutch House of Representatives, the lower house in the country’s parliament.
Plans to simplify the implementation of the EU AI Act across the EU and reduce the administrative burden for businesses were set out by the European Commission late last year. Talks between EU legislators to finalise the simplification package started earlier this spring. While the rules on ‘high-risk’ AI systems are currently due to take effect on 2 August 2026, that timeline would be pushed back – potentially to as late as December 2027 – under the proposals.