The Netherlands has made several notable advancements in AI governance. In 2019, the country launched the Strategic Action Plan for Artificial Intelligence, focusing on capitalising on societal and economic opportunities while protecting public values and human rights. To enhance oversight, the government launched a supervisory body for algorithms within the Dutch Data Protection Authority in January 2023.
In a move towards transparency, the government implemented a nationwide public sector algorithm registry in December 2022. The judiciary has also played a role, with a Dutch court ruling against the use of an opaque fraud detection algorithm (SyRI) in 2020. On the international stage, the Netherlands is actively participating in efforts to address the responsible use of AI in military applications.
The Dutch government has underscoreded the need to capitalise on AI's opportunities while safeguarding public values and fundamental rights. Its national strategy commits to principles like non-discrimination, privacy protection, and human dignity.
However, the country has faced challenges in putting these principles into practice. A major scandal erupted in 2021 when thousands of families, disproportionately of ethnic minority backgrounds, were wrongly accused of child welfare fraud by a discriminatory algorithm. This led to policy shifts including a mandatory AI registry, a new algorithmic supervisory body, and mandatory human rights risk assessments for public sector AI systems.
The Netherlands has been at the forefront of efforts to increase transparency around government use of algorithms. Amsterdam launched an AI registry in beta version in September 2020, which was used as a blueprint for the nationwide registry for public sector algorithms implemented in December 2022.
The registry aims to provide citizens with information on the purpose, legal background, application, human rights impact assessments, and oversight of algorithmic systems used by government agencies. While currently voluntary, the aim is to make it compulsory by the end of 2025.
Despite commitments to responsible AI use, the Netherlands has drawn criticism for its deployment of predictive policing systems. The nationwide Crime Anticipation System has been criticized by the Dutch Court of Auditors for its lack of oversight, privacy, and transparency. An experimental "Sensing Project" in Roermond, which used cameras and sensors to monitor all people driving in and around the city, was halted in 2022 after human rights concerns were raised.
The Netherlands has taken an active role in shaping international AI governance frameworks. It has endorsed the OECD AI Principles, UNESCO's AI ethics recommendations, and played a leading role in discussions on the responsible use of AI in military applications.
In February 2023, the country hosted an international summit on the responsible application of AI in the military domain and launched a Global Commission on Responsible AI in the Military Domain. The Netherlands is also contributing to negotiations on the Council of Europe's AI treaty.
This country report is our interpretation and summary of the "CAIDP Artificial Intelligence & Democratic Values Index 2023". The full report can be found here - https://www.caidp.org/reports/aidv-2023