Exploring the Role and Processes of NAIAEAG

National Artificial Intelligence and Algorithm Expert Advisory Group (NAIAEAG), there’s no denying that it sounds impressive, but what exactly do they do?

In this article, we'll take a closer look at NAIAEAG’s role within Health New Zealand’s (HNZ) ecosystem and how they help ensure AI technologies and advanced algorithms are developed, implemented and used safely and effectively in our health system. We’ll start by exploring their responsibilities,

with insights from Cheng Kai (CK) Jin, before examining how the team evaluates AI projects and tools.

CK started out as a doctor, spending the majority of his clinical placement at HNZ’s Waitemata District during a time of significant digital transformation. Inspired by the potential impact of data and AI in health, he pivoted to health informatics and artificial intelligence. He is now one of the technical experts with NAIAEAG, who provides oversight and advice for AI and algorithms that are being used or developed for use within HNZ or using its data.

What are the responsibilities of NAIAEAG?

The list below is not exhaustive, but is intended to outline NAIAEAG’s core responsibilities with examples and commentary to help put their work into context.  

1.   Provide advice to groups seeking to develop AI tools for use within HNZ’s environment or using its data.

NAIAEAG is made up of members who bring specialist knowledge across multiple disciplines essential to responsible AI implementation. The group has representation from clinical, ethics, privacy, data governance, legal and technical backgrounds; this multidisciplinary approach ensures that advice addresses not just the technical feasibility of proposed AI solutions, but also other considerations within HNZ's complex healthcare environment.

Many NAIAEAG members are active researchers within their roles, providing them with knowledge of the data landscape, clinical workflows, and operational challenges that AI developers must navigate. This perspective proves invaluable when advising external partners or internal teams who may not fully understand the nuances of working within New Zealand's health system.

The advisory process is designed to be flexible and supportive, with NAIAEAG reviewing proposals at various stages of the development lifecycle – from initial concept through to implementation. Early engagement is particularly encouraged, as it allows potential issues to be identified and addressed before significant resources are invested in development.

2.  Holds the authority to endorse AI tools and projects within HNZ.

NAIAEAG serves as the primary endorsing body for AI tools and projects within Health New Zealand, providing oversight to ensure that AI implementations meet standards for safety, effectiveness, and compliance. This authority is exercised through collaborative relationships with cybersecurity, privacy, and procurement teams, fostering a coordinated approach where all approving bodies maintain awareness of AI initiatives occurring across the organisation.

Rather than functioning as an approval group, NAIAEAG prioritises providing advice and considerations to project teams. The group's approach emphasises best practice guidance, working collaboratively with proposers to address concerns and improve proposals rather than endorsing or declining projects outright.

While the majority of proposals receive endorsement through this process, a limited number have been declined. These rejections have primarily occurred when implementation requests lacked sufficient evidence supporting the AI tool's effectiveness, or when supporting publications were unclear or potentially misleading about the tool's capabilities and limitations.

3.   Maintains a register of AI and algorithms and reviews the ongoing safety and effectiveness of tools.

NAIAEAG maintains a register of all AI initiatives that it has reviewed, encompassing a broad spectrum of tools ranging from simple clinical risk scores to generative AI applications. Although Health New Zealand is still early in its AI journey, maintaining awareness of AI initiatives occurring within Health NZ is crucial for ensuring appropriate oversight and governance. The register enables NAIAEAG to know what tools are being used within the organisation and to ensure that their performance and safety standards are maintained over time.

4.   Works with ethics, privacy, legal, data and commercial teams from within HNZ to manage risks associated with AI, whilst supporting research and enabling tools to be adopted by HNZ.

NAIAEAG's multidisciplinary approach enables the group to provide comprehensive advice from these diverse perspectives, ensuring that each AI proposal receives evaluation across all critical domains from the outset.

This approach means that project teams receive coordinated guidance that addresses potential concerns. For instance, when evaluating an AI tool, the privacy representative can immediately identify data protection requirements while ethics members consider patient consent implications – all within a single advisory process.

NAIAEAG maintains a strong relationship with other relevant approving bodies across Health NZ, including cybersecurity, privacy, and procurement teams. This coordination ensures that all stakeholders remain informed about AI initiatives.

5.   Set expectations with HNZ Staff around the use of AI and algorithms.

As the primary endorsing body for AI tools within Health New Zealand, NAIAEAG takes responsibility for ensuring staff understand both the capabilities and limitations of approved AI systems. Beyond simply reviewing and endorsing AI implementations, the group works to educate users about potential risks and provide guidance on safe usage practices.

A practical example of this approach can be seen in NAIAEAG's endorsement of AI scribes for clinical documentation. While endorsing this technology for use within Health New Zealand, the group specifically highlighted risks that staff needed to understand and manage. These included the risk of AI hallucinations, where the system might generate plausible but incorrect information, and automation bias, where users over-rely on AI outputs without applying appropriate clinical oversight. The group also identified potential workflow disruptions that could affect established clinical processes and provided guidance on managing these transitions effectively. By setting clear expectations upfront, NAIAEAG helps create a culture of informed AI use rather than blind adoption.

For a full list of their responsibilities and activities, see NAIAEAG’s Terms of Reference 

What is the process for evaluating projects and tools?

NAIAEAG has developed an AI checklist, which covers key questions relevant to the development, testing and implementation of AI tools within HNZ. Groups submitting proposals for advisory input or seeking endorsement for the development or implementation of tools will be required to complete this form.

Proposals are reviewed by a multidisciplinary team of experts, both internal and external to HNZ, allowing projects, research and tools to be evaluated through different lenses.

The following perspectives are considered:

·       Utility and feasibility in clinical practice 

·       Technical performance, reliability, and integration with existing systems

·       Data quality, governance, and interoperability requirements

·       Patient and population perspectives

·       Ethical principles including transparency

·       Equity, bias and discrimination

·       Māori health and Māori data sovereignty

·       Legal issues including data privacy, confidentiality and security

·       Contractual agreements, partnerships, ownership, accountability, intellectual property, monitoring and audit

·       Learning from international best practice and guidelines

This evaluation framework ensures that AI development and implementation not only meet technical requirements, but also align with Health New Zealand's commitment to safe, equitable, and culturally responsive healthcare delivery.

For more information about the AI proposals process or to request a proposal review, please contact CK.Jin@tewhatuora.govt.nz