Norway's National AI Regulatory Sandbox — Testing High-Risk and Generative AI Before Deployment
Norway
Since 2021, Norway's data protection authority has run about 20 time-boxed AI sandbox projects with public bodies — from NAV's …
New Zealand · Wellington · See the New Zealand profile
Launched 28 July 2020 by Stats NZ with 21 founding agencies, the Algorithm Charter for Aotearoa New Zealand grew to 28 signatories by its Sept 2022 independent review — which found broad support for its principles but rated implementation 'OK, could do better.'
The Algorithm Charter for Aotearoa New Zealand was launched on 28 July 2020 by then-Minister of Statistics Hon James Shaw, led by Stats NZ, as a voluntary, government-wide commitment on the use of algorithms in public decision-making. Twenty-one agencies signed at launch, including the Department of Corrections, Ministry of Education, Ministry of Justice, Inland Revenue, Oranga Tamariki and the New Zealand Defence Force.
Get signatory agencies to explain how decisions are informed by algorithms, embed a Te Ao Maori perspective, and operate algorithms 'in a fair, ethical, and transparent way.'
By the charter's independent review, published September 2022, signatories had grown to 28 agencies, including the New Zealand Police.
The review found 'almost universal support' for the charter's principles, but rated implementation 'OK, could do better.' It reported that most agencies feel there is a gap between the charter's high-level principles and concrete compliance practice; that citizens have 'very little opportunity' for individual recourse on algorithm-informed decisions; and that agencies identified 'a lack of a clear oversight body' and had resorted to self-monitoring. It recommended establishing a dedicated oversight function alongside 24 practical improvements.
As a voluntary, non-legislative charter, it remains an early, high-profile case of public-sector AI governance - but its own independent review shows the gap between signing a set of principles and having the institutional machinery to enforce them.
Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.
Where this practice's information was retrieved from, and when.
Norway
Since 2021, Norway's data protection authority has run about 20 time-boxed AI sandbox projects with public bodies — from NAV's …
United Kingdom
Consult, built by the UK's i.AI unit with the Alan Turing Institute, uses generative AI to analyse tens of thousands …
Taiwan
AI-assisted civic deliberation platform run by Taiwan's PDIS, using open-source Polis AI to map consensus on ~30 policy issues since …
Canada
Since April 2019, Canada's Directive on Automated Decision-Making has made federal departments score AI risk and publish Algorithmic Impact Assessments …
Open full copilot Grounded in cited practices — always check the sources.