|
The Netherlands Court of Audit's Recurring Algorithm Audits — a Government Oversight Model for Public-Sector AI |
Evidence: Observational / pre–post
|
Top 1% 93/100 |
Good practice |
The Hague, Netherlands |
— |
|
FRAIA — the Netherlands' Fundamental Rights and Algorithms Impact Assessment, Piloted Across 15 Government Algorithms |
Evidence: Descriptive / self-reported
|
Top 6% 87/100 |
Good practice |
The Hague, Netherlands |
— |
|
DFROG — De Nederlandsche Bank's Machine-Learning-Tested Nowcasting Model for GDP Growth |
Evidence: Quasi-experimental
|
Top 6% 87/100 |
Good practice |
Amsterdam, Netherlands |
— |
|
Woogle — the Dutch NLP Search Engine for 3 Million Freedom-of-Information Files, Now Being Tested on Estonia's Records |
Evidence: Descriptive / self-reported
|
Top 12% 80/100 |
Good practice |
Amsterdam, Netherlands |
— |
|
Amsterdam and Helsinki's Public Algorithm Register — the World's First City AI Registers |
Evidence: Observational / pre–post
|
Top 12% 80/100 |
Good practice |
Amsterdam, Netherlands |
— |
|
Overheidsbrede Monitor Generatieve AI — the Netherlands' Government-Wide Tracker of GenAI Use by Civil Servants |
Evidence: Observational / pre–post
|
Top 12% 80/100 |
Good practice |
The Hague, Netherlands |
— |
|
Pronto — Port of Rotterdam Authority's AI-Assisted Port Call Optimisation Platform |
Evidence: Observational / pre–post
|
Top 22% 73/100 |
Good practice |
Rotterdam, Netherlands |
— |
|
NCSC-NL's AI-Assisted Vulnerability Advisories — Three Times Faster Cybersecurity Guidance for the Netherlands' Critical Infrastructure Operators |
Evidence: Descriptive / self-reported
|
Top 22% 73/100 |
Good practice |
The Hague, Netherlands |
— |
|
Smart Check — Amsterdam Built a 'Fair' Welfare-Fraud AI, Then Shelved It When Bias Reappeared in Live Testing |
Evidence: Quasi-experimental
|
Top 22% 73/100 |
Good practice |
Amsterdam, Netherlands |
— |
|
Gemeente Rotterdam's Municipal Algorithm Register — 31 AI Systems Disclosed Under a Shared European City Standard |
Evidence: Observational / pre–post
|
Top 22% 73/100 |
Good practice |
Rotterdam, Netherlands |
— |
|
The Netherlands' National Algoritmeregister — Europe's Largest Public Algorithm Register, and What the Court of Audit Found Missing |
Evidence: Observational / pre–post
|
Top 32% 67/100 |
Good practice |
The Hague, Netherlands |
— |
|
Rekenkamer Utrecht's Algorithm Audit — 45 Algorithms in Use, Five Examined, One Running Without the Human Oversight Policy Requires |
Not assessed |
Top 46% 60/100 |
Good practice |
Utrecht, Netherlands |
— |
|
Rekenkamer Metropool Amsterdam's Algorithm Audit — Amsterdam's Framework Covered 49 of 61 Risks, Yet Only One Algorithm Was Tested for Fairness |
Not assessed |
Top 46% 60/100 |
Good practice |
Amsterdam, Netherlands |
— |
|
Hollandse Brug's 146-Sensor Network — Rijkswaterstaat's Real-Time Bridge Monitoring for Predictive Maintenance |
Evidence: Descriptive / self-reported
|
Top 46% 60/100 |
Good practice |
Almere, Netherlands |
— |
|
The Netherlands' National AI Regulatory Sandbox — a Single Access Point Built on a 2024 Pilot |
Evidence: Descriptive / self-reported
|
Top 46% 60/100 |
Good practice |
The Hague, Netherlands |
— |
|
Vlam-chat — the Netherlands' whole-of-government secure AI assistant for civil servants |
Evidence: Descriptive / self-reported
|
Top 46% 60/100 |
Good practice |
The Hague, Netherlands |
— |
|
ChatAmsterdam — Amsterdam's In-House Generative-AI Assistant for Municipal Civil Servants |
Evidence: Observational / pre–post
|
Top 62% 53/100 |
Good practice |
Amsterdam, Netherlands |
— |
|
UWV's Risicoscan Verwijtbare Werkloosheid — the Netherlands' Audited Unemployment Fraud-Risk Score |
Evidence: Quasi-experimental
|
Top 62% 53/100 |
Good practice |
Amsterdam, Netherlands |
— |
|
Kadastrale Kaart Next — the Netherlands Rebuilds Its National Cadastral Map With AI, Under Independent Scrutiny |
Evidence: Observational / pre–post
|
Top 62% 53/100 |
Good practice |
Apeldoorn, Netherlands |
— |
|
Voluntary Algorithm Registration in Dutch Municipalities — What a 2025 Audit of the Ten Largest Found |
Not assessed |
Top 83% 40/100 |
Good practice |
Amsterdam, Netherlands |
— |
|
DUO's Student-Grant Fraud-Check Algorithm — Dutch Risk Profiling Found Indirectly Discriminatory |
Not assessed |
Top 83% 40/100 |
Good practice |
Groningen, Netherlands |
— |
|
The Netherlands' Crime Anticipation System — A Decade of Hotspot Predictions With No Proven Crime Reduction |
Evidence: Quasi-experimental
|
Top 83% 40/100 |
Good practice |
Amsterdam, Netherlands |
— |
|
Rotterdam's Welfare Fraud Risk Model — a 315-Variable Algorithm Suspended After Audit and Investigative Scrutiny |
Not assessed |
Top 89% 33/100 |
Good practice |
Rotterdam, Netherlands |
— |
|
Gemeentelijke Chatbots Getest — an Independent Audit Finds Dutch Municipal AI Chatbots Answer Only 1 in 10 Questions Correctly |
Evidence: Observational / pre–post
|
Top 89% 33/100 |
Good practice |
The Hague, Netherlands |
— |