Top 55%
in this catalogue (354 scored practices)
Scores cluster high, so position within the catalogue is often more telling than the number alone.
Evidence of impact / measured public value1/3
Evidence is limited to a 2023 exploration phase and a 2024 pilot (40 questions, 25 organisations); the full sandbox has not yet operated, so no case-outcome or compliance-impact data exists.
Transparency, fairness & accountability2/3
The design proposal, governance model and single-access-point structure are published in full by the AP, but no operating transparency reports (e.g. published sandbox decisions) exist yet since the sandbox is not yet live.
Transferability / demonstrated replication2/3
The published design is explicitly built to be usable across all sectors via one intake point coordinated by eight authorities, and its single-access-point model is referenced in other EU states' own sandbox planning documents.
Scalability beyond pilot1/3
Designed for national multi-sector scale but not yet tested at volume — only 25 organisations engaged during the 2024 pilot that shaped the design.
Governance, capability & sustainability3/3
Governance is unusually explicit: the RDI and AP jointly coordinate, all eight AI Act market-surveillance authorities are named participants, and the mechanism has a clear EU AI Act Article 57 legal basis.
Evidoria. The Netherlands' National AI Regulatory Sandbox — a Single Access Point Built on a 2024 Pilot. Persistent ID: 0a90592c-bfc5-4019-8cba-73d42e5f2d0e.