Artificial Intelligence and Algorithm Expert Advisory Group
The Artificial Intelligence and Algorithm Expert Advisory Group guides artificial intelligence (AI) development at Health New Zealand.
About the National Artificial Intelligence and Algorithm Expert Advisory Group
The National Artificial Intelligence and Algorithm Expert Advisory Group (NAIAEAG) guides and endorses AI projects and research involving Health NZ.
The group:
- provides oversight for AI initiatives using Health NZ data
- makes sure AI projects align with ethical, technical, clinical, and operational standards
- evaluates projects through a multidisciplinary lens
- welcomes input from people with expertise or innovative ideas to support safe and effective AI use.
All AI development or implementation plans must be registered with the group.
Our members
| Name | Title |
| Dr Robyn Whittaker (Chair) | Clinical Director, Evaluation and Pathways, Data and Analytics; Planning, Funding and Outcomes — Health NZ |
| Jon Herries | Programme Director — Delivery, Digital Services — Health NZ |
| Rochelle Style | Bioethics and Health Law Consultant Independent researcher |
| Paul Muir | Clinical Service Planning — Lead, Office of Chief Medical Officer — Health NZ |
| Andrew Sporle | Deputy Director, Healthier Lives National Science Challenge Associate Professor (Honorary), Department of Statistics — Waipapa Taumata Rau — The University of Auckland |
| Juliet Rumball-Smith | Director of Intelligence, National Public Health Service — Health NZ |
| Greig Russell | Principal Advisor and Clinical Data Scientist, Haoura Māori Service — Health NZ |
| Amanda Mark | Principal Legal Counsel — Health NZ |
| Kerry Hiini | Equity Innovation Consultant — Health NZ |
| Rosie Dobson | General Manager Health Services Research and Evaluation, Evaluation and Pathways, Planning, Funding and Outcomes — Health NZ Senior Research Fellow and Health Psychologist, Faculty of Medical and Health Sciences — University of Auckland |
| Matthew Strother | Medical Oncologist and Clinical Pharmacologist — Health NZ Senior Clinical Lecturer — University of Otago |
| Angela Ballantyne | Associate Professor of Bioethics — University of Otago |
| Evo Leota-Tupou | Director Business, Product and Brand Development — Pacific Kids Learning |
| Jamie Ioane | General Practitioner — WellMed Medical |
| Suzy McKinney | Director — Mass General Brigham Data Science Office |
| Cheng Kai (CK) Jin | Clinical Director, Artificial Intelligence Laboratory, Planning, Funding and Outcomes — Health NZ |
| Riki Kyle | Senior Advisor — Knowledge and Systems, Hauora Māori Service — Health NZ |
| David McCormack | Consultant Cardiothoracic Surgeon — Health NZ Clinical Lead, Procurement and Supply Chain and Health Technology Management — Health NZ Professor (Hon), Te Huataki Waiora School of Health — University of Waikato Associate Professor of Surgery (Hon), Faculty of Medical and Health Sciences — University of Auckland |
| James Oughton | Chief Advisor for Precision Health — Ministry of Health |
AI proposals
Complete the AI proposal form if you have a proposal you want the group to review. All proposals must be developed with, or supported by, a Health NZ staff member.
If you want more information about the proposal process, or have a completed proposal form you want the group to review, email:
Dr Cheng Kai Jin, Health New Zealand Clinical Director, Artificial Intelligence Lab, Planning, Funding and Outcomes
Artificial intelligence and algorithm expert advisory group terms of reference
Health NZ National Artificial Intelligence (AI) and Algorithm Expert Advisory Group (NAIAEAG).
The terms of reference were ratified in March 2024.
These terms of reference will be reviewed annually by the committee.
The next review date is January 2025.
The purpose is to provide oversight and expert advice about the appropriateness, safety, effectiveness, ethics, legality, social and cultural licenses and ongoing improvement of any AI or mathematical algorithm research, development, projects, partnerships, contracts or implementation by employees and contractors within Health NZ.
This group will have 2 distinct functions.
- Provision of advice to groups wishing to develop or implement AI within New Zealand’s healthcare system.
- Endorsing the specific use of AI algorithms within Health NZ. NAIAEAG will provide the endorsement to the relevant approving bodies, such as Data and Digital or Privacy, for decision making.
This document defines the terms of reference for members of NAIAEAG.
Health NZ is currently developing and operating AI and algorithms. There is a need for the organisation to understand and maintain oversight of these activities to mitigate risks associated with them and ensure benefits are provided to the New Zealand population. These activities cover all services provided by the organisation whether they are for organisational, population, people and whānau or clinical needs.
The broadest definition of AI includes any activity where a machine behaves in ways that would be called intelligent if seen in a human. We consider this to include simple algorithms through to complex machine learning, natural language processing, deep learning or general purpose AI.
The scope of NAIAEAG is to provide advice and endorsement to research groups or external entities aiming to develop or implement healthcare AI for use within Health NZ. This will encompass guidance on the design, development, ethical and legal implications, and best practices around the development and implementation of AI solutions to ensure that AI can be utilised within Te Whatu Ora.
For AI projects or tools seeking approval or funding for use within Health NZ, NAIAEAG will hold authority to endorse the utilisation of said AI projects or tool. This will involve a thorough evaluation of proposed AI systems for compliance with established safety, ethical, legal and operational standards. This expert endorsement or other will be provided in writing to the appropriate decision-making body (such as data governance, investment or funding body, clinical governance, health technology assessment, ethics or research approval) making clear the assessment of risk and recommended mitigations by the group.
The responsibilities of NAIAEAG are listed below.
- Provision of review and assessment of the appropriateness, safety, effectiveness, legality, ethics, social and cultural licenses and ongoing improvement of any AI used or implemented within Health NZ.
- Provision of advice to groups wishing to develop AI using data sourced from Heath NZ’s environment or for potential use in Health NZ health services, to ensure the appropriate principles are addressed at the design stage.
- Ensures AI development and implementation are evaluated from at least the following perspectives:
- Utility and feasibility in clinical practice
- 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
- Establishes a register and regular review process for AI and algorithms in use within Health NZ.
- Identifies risks to Health NZ and its patients, population and staff that could arise from research, development and implementation of projects and provides strategies for managing identified risks.
- Sets expectations around communication with Health NZ staff around the use of AI and algorithms in the health system.
- Supports communication with our population about the use of AI in Health NZ.
- Works in partnership with appropriate Māori governance bodies.
- Works with, and provides endorsement to, formal approvals processes in Health NZ, such as for Data Governance and access, Research and ethics, Clinical Leadership and Governance.
- Works with ethics, privacy, legal, data and commercial functions to ensure these provide sufficient interface to manage risks associated with AI and algorithms.
- Obtains external expertise as required to perform these duties.
- Provides recommendations around standardisation of processes, documentation of AI and algorithm research, and development of AI in the health system.
- Directs the need for auditing and monitoring, reviews progress and outputs of any research and development projects involving AI and algorithms.
Te Tiriti o Waitangi partnership is embedded in the committee membership and in the development and prioritisation of work plans. As a Committee of a Crown entity, we will:
- Kāwanatanga align our approach with the Government’s Māori health strategy of Pae Ora and Manatū Hauora health strategies
- Tino Rangatiratanga respect the rangatiratanga of our mana whenua partner and support Māori aspirations for mana motuhake.
Meetings will be held on a monthly basis on the last week of each month.
At least 50% of the quorum are required to be in attendance for the meeting to be held.
Membership may include appointments from within Health NZ and may include external independent experts (‘external members’) for independent advice.
The chair will seek nominations from the sector and allied organisations for membership.
Membership should include, as practicable, people with skills, experience and roles in:
- health ethics
- clinical or professional leadership
- artificial Intelligence research and development
- practical experience in implementing AI and algorithms
- data management and analytics
- biostatistics
- epidemiology and public health
- Mātauranga and tikānga Māori
- the disability sectors
- data governance and Māori data sovereignty
- data privacy
- health inequities
- health economic analysis
- health research
- digital or IT service development
- consumers of health services and consumer perspectives on the use of data
- legal and contractual.
Members are appointed for a term agreed by the member and the Chair Any member may resign at any time by giving written notice to the Chair.
Any member who is absent without reasonable excuse from three consecutive meetings shall be considered to have vacated their membership
Any member may be removed from membership by Health NZ at any time and entirely at its discretion, by written notice to the member.
Unless they sooner resign, vacate or are removed from their office, every member shall continue in office until their successor comes into office.
The responsibilities of the Secretariat include the following.
- Circulate draft minutes and actions within 2 weeks of the meeting.
- Circulate agenda items at least 5 working days ahead of each meeting to allow for members to raise conflicts of interest.
- Circulate agenda pack at least 3 working days ahead of each meeting to allow for reading time. For members with conflicts of interest, specific agenda documents will not be provided for the specific item.
- Maintain a Conflicts and Declaration of Interests Register.
- Maintain an AI register of AI tools used in Health NZ.
- Maintain a current action list.
The responsibilities of Members include the following.
- Members of the group must perform their functions in good faith, honestly and impartially, and avoid situations that might compromise their integrity or otherwise lead to conflicts of interest.
- Members are expected to be punctual, polite, professional, engage in intellectual discourse and be respectful of others’ views.
- Agenda items to be submitted to the Secretariat at least 5 working days prior to each meeting.
- Members with conflicts of interests will need to be raised at least 3 working days ahead of each meeting to ensure non-disclosure of information to conflicted members.
- Apologies to be sent as soon as possible to the Secretariat for quorum purposes and if required, a suitably briefed and authoritative delegate to be in attendance with the approval of the chairperson.
- Members are able to provide their input via email prior to meetings where they are unable to attend.
- Members are invited to submit edits to draft minutes, which will be finalised at the start of each meeting. Undertake additional activities agreed by the group (such as commenting on advice or guidance and providing research material).
- Lead/facilitate the completion of respectively owned action items within the agreed timeframes.
- Sign a confidentiality agreement.
- Exercise all due professional care and diligence in the performance of their obligations under these Terms of Reference in accordance with the standards of skill, care, and diligence normally practised by suitably qualified and experienced persons in performing services of a similar nature.
- Be familiar with NAIAEAG’s Interest Recording and Conflict of Interest Policy and declaring all Interests which may create an actual, potential or perceive conflict of interest with their duties and responsibilities as NAIAEAG members.
Interested Health NZ staff are welcome to attend meetings any time and to bring projects to the group but they are not counted for the purposes of establishing a quorum. If they are presenting, they must declare any interests or conflicts of interest they have in accordance with NAIAEAG’s Interest Reporting and Conflicts of Interest Policy.
Submission process
The process for submitting proposals to NAIAEAG will differ based on whether the proposal is seeking advisory input or seeking endorsement for development or implementation within Health NZ.
Proposals intending to seek advice will be submitted using a standardized template to the Secretariat.
For proposals seeking endorsement for algorithm development or use within Health NZ, a more comprehensive checklist is required. This checklist is designed to cover all aspects pertinent to the development and deployment of algorithms within a healthcare setting, ensuring that proposals meet the necessary standards and regulations.
Initially, submissions will be screened by the NAIAEAG Secretariat and prioritised for discussion based upon organisational or clinical need and urgency.
It is expected that all proposals will have a Health NZ partner or sponsor, who will be asked to be part of the presentation, such as on behalf of a clinical or operational service that is supporting the proposal.
Decision-making
The group will strive for decision-making based on consensus for all proposal submissions. A consensus is defined as an agreement reached by 50% or more of the members present. All members will have the ability to contribute to the consensus.
Members who are absent from the meeting will not be able to contribute to the consensus.
All decisions should first be attempted to be resolved through discussion and mutual agreement.
In cases where the vote is neutral, the Chair will have the final decision. In instances where consensus or vote is not achieved, the nature of the disagreement and the reasons provided by the dissenting members will be documented.
The minutes of the meeting will clearly reflect the discussion, including all viewpoints expressed, and the rationale behind the final decision or the lack of consensus.
The group will provide advice directly to those requesting it either in a meeting or in follow-up communications.
The group will provide endorsement and recommendations to the Te Whatu Ora sponsor or partner and to the relevant decision-making body in writing. The group will report regularly on their reviews, findings and endorsements (or otherwise) to the National Directors of Data and Digital / Service Improvement and Innovation and others as appropriate.
Individual members or those requesting endorsement are able to escalate any issues or disagreement to the Te Whatu Ora National Innovation Network for further review.
Members must disclose all Standing Interests and Interests in a Matter along with any actual, potential or perceived conflicts of interest in accordance with NAIAEAG’s Interest Reporting and Conflicts of Interest Policy as soon as they are aware of them.
If, upon receipt of the Agenda Outline for a meeting (to be sent by the Secretariat 5 working days ahead of each meeting), a member becomes aware that they have an Interest in a Matter in relation to that Agenda, they must disclose that Interest(s) to NAIAEAG’s Chair at least 3 working days prior to each meeting.
Upon receipt of a member’s disclosure of an Interest in a Matter in relation to an Agenda item, the Secretariat will refrain from providing any documentation about the matter to that member prior to the meeting at which time NAIAEAG will decide, in accordance with the Interest Reporting and Conflicts of Interest Policy: (1) whether or not the Interest creates a conflict of interest; and (2) if so, how that conflict will be managed.
All disclosures of an Interest in a Matter in a meeting, whether it constitutes a conflict of interest and, if so, how it was managed, must be recorded in the Minutes of that meeting in accordance with the Interest Reporting and Conflicts of Interest Policy.
All business of the NAIAEAG, other than information before the NAIAEAG already available in the public domain or intended for dissemination in the public domain, is confidential and must be treated as confidential by all group members.
Members are not to disclose any confidential information to anyone outside the group, other than to group members, without the prior approval of the Chair in consultation as appropriate with the group member (unless the matter concerns the conduct or performance of the Chair) and are to treat this material with the utmost care and discretion.
Advice, recommendations and endorsements will be reported to the National Directors Data and Digital / Service Improvement and Innovation.
A brief summary of advice, recommendations or approvals will be made publicly available.
A National Register of AI in use within Health NZ will be made publicly available.
Independent Members are not liable to Health NZ for any act or omission done or omitted in his or her capacity as an Independent Member, provided that the Independent Member acted in good faith, and with reasonable care, in pursuance of these terms of reference.
Independent Members are indemnified by Health NZ:
- for costs and damages for any civil liability arising from any action brought by a third party in respect of any act or omission done or omitted in the Independent Member’s capacity as an Independent Member of the NAIAEAG, provided that the Independent Member acted in good faith, and with reasonable care, in pursuance of these terms of reference
- for legal and other expenses arising from any successfully defended criminal proceeding in relation to any such act or omission.
The Independent Member may engage legal counsel of the Independent Member’s own choosing in relation to any civil and/or criminal proceedings brought against the Independent Member in respect of any act or omission done or omitted in the Independent Member capacity as a member of the NAIAEAG.
Fees paid to members are determined in accordance with the Cabinet Fees Framework for members appointed to bodies in which the Crown has an interest. The Expert Advisory group is classified among 'Group 4: All other committees and other bodies.'
Fees are paid for preparation, attendance and travel in relation to scheduled committee and working group meetings. Fees are not paid to salaried public servants.
Fees payable to members per meeting are a:
- meeting attendance fee of $560 including GST for any meeting of between 6 and 8 hours, for shorter meetings, a fee of $560/8 per hour
- meeting preparation fee of $280 including GST
- travel fee of $280 including GST for travel of more than 3 hours.
Fees are categorised as honoraria and as such are subject to withholding tax pursuant to Schedule 4 Part B of the Income Tax Act 2007 No 97.
Heath NZ will reimburse members for actual and reasonable expenses in attending meetings. Any accommodation and travel must be arranged by Health NZ. Members may submit a travel expense claim for costs such as parking.