The Dutch Data Protection Authority (AP) has launched a new directorate for the supervision and coordination of AI and algorithms. Among other things, the directorate is preparing for future market supervision of the AI Act and is working with the Digital Infrastructure Inspectorate (RDI) on an AI sandbox and a joint coordination centre.
The AP has not yet been formally designated as the market supervisory authority. This requires the Dutch Implementing Act for the AI Act. The AP will only have powers for market supervision once that Act comes into force. However, the supervisory authority is already working on cooperation, information and support for responsible innovation, and is building up knowledge of, among other things, AI safety.
The new directorate consists of the AI/Algorithms Risk Analysis & Coordination (ARC) and AI Market Supervision (MTA) departments. ARC investigates the effects of AI and algorithms on people, fundamental rights and society. Among other things, the department focuses on high-impact algorithmic systems, algorithm registers and auditing frameworks. MTA is developing knowledge and technical expertise to investigate and test AI systems and is preparing supervision of the AI Act.
The AP and the RDI are also jointly setting up an AI sandbox. In this controlled environment, organisations can develop innovative AI systems together with supervisory authorities and bring them into line with the AI Act. According to the AP, rules for AI already apply: some applications are prohibited, and transparency requirements apply to, among other things, AI-generated videos, images and texts. From 2 December, there will also be a ban on AI that virtually undresses people without their consent. In the Netherlands, work is still under way on the Implementing Act for the AI Act.
