FRAIA — the Netherlands' Fundamental Rights and Algorithms Impact Assessment, Piloted Across 15 Government Algorithms

FRAIA (Fundamental Rights and Algorithms Impact Assessment) is a structured tool for identifying how a specific algorithm may affect fundamental rights, originally developed by Utrecht University in 2021 for the Ministry of the Interior and Kingdom Relations. It is meant to be applied before a government organisation develops, procures or deploys an AI or algorithmic system.

In 2023 the Ministry commissioned Utrecht University and Rijks ICT Gilde (the national ICT guild) to run a structured pilot programme to build experience with FRAIA and encourage its adoption across government.

Over Fall 2023 to Spring 2024, fifteen pilots were carried out with different government organisations — split evenly, eight supervised by Utrecht University and eight by Rijks ICT Gilde — covering algorithms at every stage of maturity, from not-yet-built to already in production, and spanning both machine-learning systems and simpler rules-based ones.

The published pilot report found that participants generally came away with a positive impression despite initial scepticism, valuing the structured discussion FRAIA generated across legal, ethical and technical perspectives. The main documented drawback was that the process is time-consuming. The pilots also surfaced a need to train staff internally in algorithmic-impact thinking and to integrate the assessment more tightly into existing laws and procurement rules.

Following the pilots, the Dutch House of Representatives endorsed making FRAIA use mandatory for government algorithms, moving the tool from a voluntary good-practice instrument toward a binding governance requirement — though as of the pilot report this was an endorsement, not yet a fully enforced legal mandate. The evidence here is genuinely about process and institutional uptake rather than a measured reduction in algorithmic harm.