Heartline Connect Auckland Meetup: Anca Yallop and Crystal Yap on the Ethics of AI in Youth Mental Healthcare
The first Heartline Connect meetup was held on Friday 11 September at Olea Cafe and Eatery in East Tāmaki, Auckland. The room was made up of practising clinicians and therapists, along with people training towards clinical psychology. Heartline Connect is a clinician community of the AI Health Innovation Hub (AHIH), presented by EmoEx.
The afternoon centred on a talk from Associate Professor Anca Yallop and Professor Crystal Yap of Auckland University of Technology, who presented their Health Research Council-funded study into the ethics of AI-driven technologies in youth mental healthcare. Will Zhang then demonstrated a tool EmoEx is building with clinicians and took feedback from the room. The rest of the afternoon was drinks, nibbles and conversation.
The Ethics of AI-Driven Technologies in Youth Mental Healthcare
Associate Professor Anca Yallop, Auckland University of Technology Business School
Professor Crystal Yap, Auckland University of Technology Business School
Anca opened by setting out the shape of the project. Funded by the Health Research Council, it started in October 2025 and runs for eighteen months.
"We are very interested in your views on how and what we can do better in terms of using AI technologies in youth mental healthcare."
Background
Anca traced the project back to a shift in government priorities. It became clear that the health sector was going to be encouraged to explore AI, and in May 2025 the Health Research Council put out a call for proposals on AI in healthcare. The framing at the time was productivity, alongside a recognition that the tools have to be safe and used responsibly.
She then set out the regulatory position. New Zealand is taking a ‘light touch’ approach. There is no dedicated AI Act. The country is leaning on existing legislation, including the Privacy Act and the Human Rights Act, with the rest handled through voluntary guidance. She contrasted this with the European Union, where the EU AI Act is being implemented in stages out towards 2028, noting that the learning is still going on there too.
The national guidance rests on a set of principles aligned with the OECD AI Principles: AI should be accessible. It should be safe and effective. It should provide good value. It should be accountable, which Anca described as working in a transparent and well-controlled manner so that patients are protected. And it should be equitable, which means the benefits have to reach a wide range of New Zealanders, particularly those with higher health needs, including Māori and Pacific peoples.
The guidance, as she described it, is focused on accelerating private sector adoption and innovation.
The Research Team
Anca and Crystal both work in the Marketing and International Business department at the AUT Business School. Anca comes to the project from technology and data ethics, and Crystal from consumer wellbeing and digital wellbeing. The wider team brings in ethics, Māori health, AI and neuroscience, clinical psychology, and youth mental health and wellbeing, drawn from AUT, the University of Auckland and clinical practice.
Objectives
The project has three objectives:
The first is to understand what is happening now, and what a gold standard would look like for the ethical, responsible and effective implementation of AI in Aotearoa New Zealand's youth mental healthcare system.
The second is to talk to practitioners. Not only those already using AI in clinical practice, but those considering it. The interest is in their experiences and perspectives, and specifically in the ethical tensions they perceive when these tools enter the work.
The third is to turn that into something usable. An evidence-based policy framework, co-designed with community partners, producing clear guidelines, principles and recommendations for responsible use.
Anca was direct about why the third objective needs the first two. There is no existing framework to work from, other than the EU AI Act.
Research Methodology
Crystal took the second half of the talk and spoke about the methodology, framing it as the element that determines whether the project is meaningful.
The team is using community-based participatory research. As Crystal described it:
"We are here to work with the community. Not to just work on the community."
She set out what that means in practice. The first is shared decision making. The research team does not decide alone how a study is designed or how it handles ethics. Community members are brought in early and consulted on those choices rather than presented with them.
The second is mutual benefit. Crystal was explicit that publication is not the measure of success. The question is what the community gets from the work, including patients, and whether it helps practitioners carry out their jobs.
The third is checked interpretation. When the qualitative insight comes in, the team does not want to sit alone with it and decide what it means. The intention is to take findings back to practitioners and ask why. Crystal used the example of an insight the team had never anticipated, and said the useful move is to return it to the people it came from rather than explain it from the outside.
The fourth is attention to power. Crystal noted that in most fieldwork the interviewer holds a perceived authority in the process, and said the team is working towards a balance rather than treating that as given.
The study began with a scoping stage covering both the global and the New Zealand picture. Anca added detail here: the team looked not only at academic literature but at policy, to understand the landscape politically as well as academically, and to see what practitioners in other parts of the world are doing. Crystal made the point that scale and sophistication elsewhere do not settle the question locally. There are apps operating in the United States well ahead of anything available here, and whether they suit the New Zealand context is a separate matter.
That stage is complete. The project is now in its empirical phase, which is the part that needs New Zealand voices.
Crystal closed on dissemination. The intention is not to stop at a finding. The team wants to present the work to organisations directly once the insight is in.
Anca returned to the same point from a different angle, and it served as the close of the talk.
"It is very easy to just go with the AI buzz. But we need to be mindful of how we do that."
Taking Part
Anca and Crystal are recruiting for the empirical stage of the study. They are looking for clinical practitioners, including clinical psychologists and psychiatrists, for interviews. For focus groups, they want a wider range of perspectives, including technology experts, social workers, policymakers, ethicists, youth organisations and advisers.
For more information and how to get involved, you can read more about the study here.
EmoEx Clinicians’ Dashboard
Following Anca and Crystal’s presentation, Will Zhang showed the room a tool EmoEx is building with clinicians. While not live, the purpose of showing the prototype at this stage was to gather feedback and to co-design it with the people who would use it.
About Heartline Connect
Heartline Connect is a community of clinicians and therapists under AHIH. It is the destination for EmoEx companion users who need more than an AI can offer them, and it is the group EmoEx is developing clinician-facing tools alongside.
It is free to join, and there is no commission. Referrals begin at the end of September 2026 and come through EmoEx's companions.
If you are a clinician or therapist and want to know more about Heartline Connect, you can find out more here.
Comments