AI Adoption in Healthcare in New Zealand: A Living Whitepaper

Introduction

  • New Zealand’s healthcare system faces mounting pressures (aging population, staffing shortages (techalliance.nz)) that have prompted leaders to seek technology solutions. AI is seen as a way to augment scarce clinical resources and improve efficiency. For example, the AI Forum NZ notes that chronic nurse and doctor shortages mean “AI is the answer” to amplify limited expertise (e.g. automating diagnostics) (techalliance.nz).
  • The government emphasizes a “precision health” approach: tailoring care to individuals using data (genes, lifestyle) and safe AI tools (www.health.govt.nz) (www.health.govt.nz). New Zealand is still in the “early stages” of adopting AI in health, focusing on small pilots and non-clinical uses to improve access and quality (www.health.govt.nz) (www.hinz.org.nz).
  • A cautious-but-forward strategy is clear: Te Whatu Ora has explicitly warned staff not to use unvetted generative AI (like ChatGPT) for patient care in 2023 (www.hinz.org.nz), even while new AI pilots and research funding appear. Overall, stakeholders stress that AI in NZ healthcare must be safe, equitable and transparent (www.health.govt.nz) (www.hinz.org.nz).

Recent Developments

  • Te Whatu Ora guidance (Sept 2023): National health service issued an official caution against the use of large language models and generative AI for clinical care. Staff must not feed patient data into ChatGPT/LLMs or rely on them for medical decisions, since these tools “have not been validated as safe and effective” (www.hinz.org.nz). This precaution reflects concerns over privacy, bias and reliability (www.hinz.org.nz).
  • Population health AI pilot (Feb 2025): Health Minister launched an AI-assisted diabetic retinal screening pilot in Māngere (South Auckland) (www.beehive.govt.nz). Community technicians capture retinal images, and an AI tool pre-screens them (with ophthalmologist backup). This enables real-time grading and faster referral of at-risk patients, significantly increasing screening volumes and reducing specialist wait times (www.beehive.govt.nz) (www.beehive.govt.nz).
  • Radiology AI (May 2025): Health New Zealand’s AI Lab is collaborating with radiologists to explore AI for high-volume imaging (e.g. chest X-rays, fractures, CT scans) (www.hinz.org.nz). Pilot projects aim to use AI as a “co-pilot” to reduce error rates and triage critical cases, not to replace clinicians (www.hinz.org.nz) (www.hinz.org.nz). For example, Canterbury DHB is integrating AI flagging to prioritize suspicious findings (some tools already detect 100+ abnormalities) (www.hinz.org.nz) (www.hinz.org.nz).
  • Clinical documentation AI (Apr 2025): New Zealand health-tech firm Medow Health released an AI “co-pilot” for hospital specialists (www.hinz.org.nz). The system listens during consultations, drafts patient notes and reports, and frees doctors from manual transcription. Early adopters report doctors reclaim 2–3 hours per day/specialist, allowing 1–2 extra patient consults daily (channellife.co.nz) (www.hinz.org.nz). This addresses long specialist waitlists (over 65,000 NZ patients waited months for a first appointment) (channellife.co.nz) (www.hinz.org.nz).
  • Healthcare chatbot (Apr 2024): In mental health, the No Duff Trust (a veterans’ charity) created an AI chatbot trained on 9 years of peer-support conversations (www.rnz.co.nz). This service aims to offer 24/7 assistance to veterans in crisis, under clinician supervision. The COVID-era Dartmouth study noted that well-trained chatbots can rival human therapists (www.rnz.co.nz), highlighting potential in underserved areas.
  • Industry & Events: Auckland hosts major digital health forums (e.g. HiNZ’s Digital Health Week 2023 & the forthcoming 2025 AI Summit) emphasizing AI’s role in transforming NZ healthcare (www.scoop.co.nz) (www.hrc.govt.nz). Professional bodies are issuing guidance: in Jan 2024, RANZCR (radiologists’ college) collaborated internationally on best practices for safe AI in radiology (www.nzdoctor.co.nz) (www.nzdoctor.co.nz). These discussions signal readiness to adopt AI, tempered by quality and safety checks.

Research and Initiatives

  • Government funding: In Feb 2025 the Health Research Council announced a $5 million RFP for AI in health (www.hinz.org.nz) (www.hrc.govt.nz). It seeks investigator-led projects aligned with national health priorities (2024–27), explicitly on safe, ethical AI to improve outcomes (www.hinz.org.nz) (www.hinz.org.nz). Areas of interest include streamlining care (reducing wait times, boosting efficiency) and enhancing diagnostics/predictions (www.hinz.org.nz). Grants range from $100k to $700k (up to 2 years duration) with August 2025 start (www.hinz.org.nz). Collaborative, cross-disciplinary proposals are encouraged to translate research into practice (www.hinz.org.nz).
  • Coordination networks: Following the PM Chief Science Advisor’s 2023 AI in Health report, an “AI in Health Research Network” has been formed to unite experts across NZ (www.ai-health-research.org.nz) (www.ai-health-research.org.nz). Its goals are to share knowledge, co-develop AI tools with Tiriti-aligned ethics, and create pathways from lab to hospital. Such initiatives aim to streamline efforts and promulgate best-practice AI deployments in the under-pressure health sector (www.ai-health-research.org.nz) (www.ai-health-research.org.nz).
  • Policy context: The Ministry of Health’s precision health strategy highlights AI as a future priority for diagnosis, treatment planning, workforce support and administration (www.health.govt.nz) (www.health.govt.nz). The Ministry and Health NZ are working with international partners and advisory groups to develop regulatory frameworks and oversight before scaling new AI tools (www.health.govt.nz) (www.health.govt.nz). Public reports emphasize AI’s promise (e.g. personalized preventive care) but stress caution regarding privacy, bias and equity (www.1news.co.nz) (www.health.govt.nz).

Case Studies

  • Medow Health AI assistant: Founded by Kiwi entrepreneurs, Medow Health provides an AI co-pilot for specialist doctors (www.hinz.org.nz). In practice, it records clinician–patient conversations and drafts consult notes in real-time. Early users report dramatic time savings: doctors gain 2–3 hours/day, enabling more patients per clinic and reducing burnout (www.hinz.org.nz) (channellife.co.nz). The company has deployed the tool in NZ and Australia (60k consultations/quarter) and expects rapid growth. Medow’s approach exemplifies how AI can augment clinicians without replacing them (www.hinz.org.nz) (channellife.co.nz).
  • AI Diabetes Retinal Screening (Health NZ pilot): The South Auckland pilot (launched Feb 2025) equips seven GP practices with retinal cameras. Community techs capture images, and AI software automatically grades them for disease risk (www.beehive.govt.nz). Images flagged as abnormal are diverted to specialists instantly. This “real-time” AI grading greatly increases daily screening volume, cuts wait times, and frees hospital ophthalmologists to focus on confirmed cases (www.beehive.govt.nz) (www.beehive.govt.nz). It is one of the first nationwide uses of AI in disease screening.
  • Veterans’ AI chat therapy (No Duff): Addressing gaps in mental health support, No Duff Trust built a custom AI chatbot trained on 9 years of veteran support dialogs (www.rnz.co.nz). It offers mindfulness and crisis support modules under a clinician review process. Though still in testing, this grass-roots initiative highlights AI’s role in extending 24/7 help where human resources are limited (www.rnz.co.nz) (www.rnz.co.nz). Early feedback from veterans suggests technology solutions are welcomed when formal care is lacking.
  • Global AI tools in NZ practice: Some proven technologies are already in use. For example, ambient speech recognition (Nuance’s Dragon Medical One) is approved in NZ DHBs for dictating notes (soundbusiness.co.nz), dramatically speeding up documentation (pilots overseas found dozens of clinical hours saved). AI-powered pathology analysis (e.g. computer-aided detection in CT) is also present in some hospitals (www.hinz.org.nz) (www.nzdoctor.co.nz), though specifics are often bundled into routine practice. These demonstrate low-risk “narrow AI” being adopted quietly under existing medical device regulations.
  • Cautious expansion: A 2025 AI Forum survey found ~65% of NZ health/social services organisations report some AI use, mostly in administrative, back-office roles (www.hinz.org.nz). Clinical adoption lags due to privacy and safety concerns. Health leaders assert that AI applications in NZ will follow robust validation: for instance, AI is “already approved” in areas like radiology image analysis (www.hinz.org.nz), but untested new tools (especially generative AI) are held at bay (www.hinz.org.nz) (www.hinz.org.nz). The emphasis is on augmenting – not replacing – clinicians (www.hinz.org.nz) (techalliance.nz).
  • Efficiency and workflow: Workforce shortages (∼8,000 vacancies) continue to drive interest in productivity gains (soundbusiness.co.nz). AI transcription, automated coding and triage systems (already piloted overseas) are predicted to spread. New Zealand companies (like Medow) and overseas partnerships aim to deploy these ASAP. The just-launched AI screening pilot and NZTech Forum stance reflect a trend toward AI tackling bottlenecks (screening waitlists, specialist reporting, triage) before moving to complex diagnosis (www.beehive.govt.nz) (www.hinz.org.nz).
  • Regulation and ethics: The government’s “guardrail” approach is set to continue. Frameworks for ethical AI use (e.g. Te Tiriti-informed principles (www.ai-health-research.org.nz)) will guide progress. NZ’s ACC system covers patient harm, but liability in AI errors remains unclear (brightstar.co.nz). Legal and professional bodies are examining risks (RANZCR/Govt advisories) and training needs for safe AI use (www.nzdoctor.co.nz) (brightstar.co.nz). Public trust is a priority – any AI tool deployed in clinics must be transparent about capabilities and limitations (www.hinz.org.nz) (www.hinz.org.nz).
  • Future directions: Momentum is growing for precision health: combining genomics, wearables and AI for preventive care (www.1news.co.nz). New Zealand’s newborn sequencing and genetic services lay groundwork, and experts envision AI-driven personalized screening and treatment plans. However, researchers caution on environmental impact and equity (www.1news.co.nz). AI is also expected to support underserved communities (through telehealth or targeted apps), provided access is improved. Overall, forecasts see steady integration: government, industry and academia agree that the status quo is unsustainable, so AI will be one pillar of future strategy (www.hinz.org.nz) (www.nzdoctor.co.nz).

Conclusion

  • AI is progressively entering New Zealand’s healthcare system via targeted projects and tools. Recent initiatives – from a new retinal screening AI pilot to AI note-taking systems – illustrate tangible benefits in reducing workload and wait times (www.beehive.govt.nz) (www.hinz.org.nz).
  • The consensus is that AI will complement NZ clinicians and help manage rising demand, but it must do so under strict ethical oversight (www.hinz.org.nz) (www.hinz.org.nz). Health policymakers and providers emphasize safe, equitable deployment in line with national priorities (www.health.govt.nz) (www.hinz.org.nz).
  • Continued investment (e.g. HRC funding, research networks) and collaboration (industry, academia, iwi/Māori stakeholders) are paving the way. As AI pilots mature into broader programs, this Living Whitepaper will track developments, aiming to guide stakeholders on leveraging AI for better health outcomes in Aotearoa.

Sources: Government releases, industry news, and sector reports (Health Ministry, HRC, HINZ, RNZ, NZ Doctor, AI Forum NZ, etc.) (www.beehive.govt.nz) (www.hinz.org.nz) (www.hinz.org.nz) (www.hinz.org.nz) (www.hinz.org.nz) (www.rnz.co.nz) (www.nzdoctor.co.nz), reflecting the latest information on AI in New Zealand’s healthcare.