1. A public register of government AI systems’ function, performance, and impacts
1. A public register of government AI systems’ function, performance, and impacts
We want the government to make a list— a 'register'— of the ways it uses AI. This will help us easily find out whether AI is being used in ways that impact us.
This list should include:
How AI is being used,
How well it's performing,
How this could impact our public services and our lives.
For example, the Ministry of Social Development is using relatively simple rules-based AI to check if someone applying for a student loan is already on the benefit or not. This particular system is 99.9% accurate, and any potential errors can be redirected from the automatic system to a human for review.
Another example is that Health New Zealand has endorsed the use of "ambient AI scribes" like Heidi Health, allowing staff to automatically transcribe conversations with patients. This system been shown to reduce documentation time by 10 minutes per patient, and reduces after-shift admin by 81%. However, it can still produce errors, which is why the outputs are considered draft only and must be verified by a medical professional before they can be saved into a patient's medical records.
We are asking for all of this information to be made available and in one place, making it easy for you to find and use. A register could also make it easier to understand the information, because information is presented in the same format each time.
If you want to know about which government agencies are using AI, right now you have to rely on them to voluntarily disclose this, or for someone (like a journalist) to file an Official Information Act (OIA) request and publish their findings.
Many AI systems are already in use, including Automated Decision Making systems. However, it's really difficult to find information because there is very little information published about it.
We can't ask questions if we don't know about it - transparency is the first step to trustworthiness. A public register forces proactive disclosure, and it ensures that transparency is part of the system design... not an afterthought triggered by an OIA request or a legal challenge.
We also think that a list of AI systems (functions) isn't enough - we need to see that testing has demonstrated how well these systems work (performance), and that agencies have identified how these systems will affect people (impacts). This aligns with our next idea around impact assessment.
Many countries have some type of public register, or requirement to share information about AI use in the public sector. For example:
The United Kingdom recently launched the Algorithmic Transparency Recording Standard Hub
Canada has an AI Register for government AI systems (see here for an introduction)
The Netherlands has an Algorithm Register (see here for a list of all algorithms)
Scotland has a Register of AI Systems in the public sector
The European Union AI Act will require a EU database for high-risk AI systems in late 2027
Some New Zealand public sector agencies do publish this information already, but is not found in one place. For example:
New Zealand Police publishes a Technology Capabilities List
The Ministry of Justice has a list on their Algorithm Charter page
The Ministry of Social Development describes their processes for Automated Decision Making, although they do not state when this was last updated
2019: The University of Otago's AI and Law in New Zealand Project* called for a public register of predictive algorithms
2020: Dr Karaitiana Taiuru's report on Treaty of Waitangi/Te Tiriti and Māori Ethics Guidelines for: AI, Algorithms, Data and IOT notes the necessity of transparency for informed consent when using Māori data.
2021: Taylor Fry’s Year 1 Review of the NZ Algorithm Charter (2021) recommends NZ ‘develop an annually updated register of algorithms covered by the Charter’ (see e.g. their summary on p33)
2024 March: The ARC Centre of Excellence for Automated Decision-Making and Society released a report mapping and analysing the use of ADM systems by state and local governments in Australia; they suggest that building a register would help to effect ‘standardisation of key terminology’ across government departments.
2024 June: The Public Health Communication Centre Aotearoa has an opinion piece* arguing for a register
2025: Te Kāhui Raraunga released the Māori AI Governance framework, recommending the establishment of an open and transparent register of algorithms in use by the government that is publicly available to view.
2026 Jan: The NZ Public Service AI Work Programme outlines plans to create a "Central repository of AI registries, patterns and tools", although this seems to be a non-public registry to enable public servants to share AI tools between agencies; the main purpose is to drive AI usage. This work programme also mentions "transparency" under "Safe and Responsible AI," but transparency isn't actually explained when you dive into the detail
2026 March: Johniel Bocacao describes what an all-of-government AI register could look include.
2026 May: A report* from NZ’s Maxim Institute recommends creation of an AI register for the NZ public service (see p25)
2026 May: Dr Alexandra Andhov (University of Auckland) wrote an op-ed in Newsroom that explores the cost of adopting AI in our public service, including other costs like sending money and information overseas while removing the expertise from government to ensure AI is implemented well
The details of what is included in the Register will need to be debated before they reach their final form—that's why we have asked for public deliberation. But we have some starting ideas:
Objective: What problem is the AI system trying to solve?
Capability: How does the system actually work? This could contain both the process elements (how the AI system uses data) and the technical elements (what type of AI it is).
Provenance: What datasets were used to train this system, and where did the system itself come from? How confident are we that the system can reflect New Zealand's unique demographic and cultural context?
Human-in-the-loop: At what stages do humans review the machine outputs, and can they intervene?
Performance: Has the system been tested, and how often is it reviewed for accuracy and acceptable levels of error? This could include benchmarks relevant to New Zealand and our values (for example, see Te Hiku Media's benchmarks that were designed for their purpose of language revitalisation, in contrast to generic accuracy benchmarks used overseas.)
Accountability: Who is specifically held accountable if the system produces discriminatory or flawed outcomes?
Impact: Who does this system affect (positively or negatively)? Is there a link to the impact assessment?
Not necessarily. We are open to a framework that has the flexibility to be low-admin for low-risk uses of AI. For example, a low-risk system might not need to be reviewed as often. This keeps the paperwork at a minimum, while ensuring that higher-risk uses get the planning and testing they deserve.
Additionally, in most cases, the information we are asking to be published is already held by the agencies when they design and procure these systems. There is minimal new work; we are just asking for the work that is (hopefully) already being done to be published in a way that's easy to find, and easy to understand.
Even then, narrow exceptions consistent with existing law would be required, for example systems that are used for national security, or where release of the information would prejudice the maintenance of the law or endanger public safety. Commercial sensitivity should not be a valid reason to withhold the existence of a tool, noting that no source code or proprietary secrets need to be published.
Innovation flows faster when there is certainty. Deploying AI in secret and then having it exposed later leads to public backlash and legal challenges. This slows down the pace of government innovation and adoption far more than a small amount of paperwork to publish information that already exists onto a centralised public register.
Sharing information about the expected benefits and the risks of an AI system increases trust, which increases acceptance of AI systems (see page 60 of the 2025 report on Trust, Attitudes, and Use of AI). When we have transparency, the New Zealand government can progress with the confidence it needs to adopt AI, knowing that the bumps in the road have been avoided before they became an issue.