Artificial intelligence in radiology is no longer a future idea. The question for Australian and New Zealand radiologists is simpler, what is working now, and what should we plan for next.

What is already in use

• South Australia’s public system is live. South Australian Medical Imaging has implemented AI assistance for chest X-rays across six metropolitan and four regional hospitals. The software highlights areas of interest and suggests potential diagnoses, and radiologists retain full responsibility for the report. Public coverage in January 2025 confirms the rollout and the approach.
• National-scale use in private practice. I-MED has deployed Annalise CXR at enterprise scale across Australia, supporting large volumes of chest X-ray cases each month. This is decision support for radiologists, not automation of reporting.
  

These examples show the practical value today, quicker triage of routine cases, earlier flagging of critical findings, and support for teams under workload pressure. 

What is still emerging

Complex CT and MRI studies still rely on clinical judgement, correlation with history and other imaging, and local protocols. AI can help surface patterns and prioritise worklists; however, it does not replace a radiologist’s interpretation. Professional bodies across our region continue to emphasise safe, supervised use and robust evaluation before clinical adoption. 

Compliance and safety anchors you should know

• AHPRA guidance. AHPRA has issued practical guidance that ties AI use to existing professional obligations, including safety, informed consent, and accountability for clinical decisions. If you use AI, you remain responsible for the outcome.
• TGA regulation. Many imaging AI tools are regulated as software-based medical devices. The TGA’s guidance explains how these products are classified and assessed, and what sponsors must do to comply. This matters for procurement and governance in your service. 

The New Zealand picture

New Zealand is building the foundations that support safe adoption. Te Whatu Ora has a national advisory group for AI and algorithms, and the Prime Minister’s Chief Science Advisor has documented the governance approach. There is also fresh, funded research to evaluate AI for chest X-ray reporting in public hospitals, with a two-year programme that started in 2025. For many NZ services, the focus is readiness, governance and evaluation before wider rollout. 

Reality versus hype

Real today

• Decision support in chest X-ray interpretation, including worklist triage and flagging of time-sensitive findings.
• Enterprise deployments that save time on high-volume studies while keeping the radiologist in charge.
• Clear professional and regulatory expectations for safe use.
  

Hype to ignore

• Full automation of reporting. No credible body in ANZ supports this, and current statements emphasise oversight and accountability.
• Compliance shortcuts. AI may improve efficiency; however, it adds governance requirements rather than removing them. 

What to do next

1. Map your workflows. Identify where AI can support speed and quality, such as chest X-ray triage or quality assurance, and document how responsibility stays with the reporting radiologist.
2. Check regulatory status. Confirm TGA registration or listing, vendor evidence, and local validation before deployment. 
3. Align with professional guidance. Ensure policies reflect AHPRA guidance and RANZCR’s position on safe, supervised use. 
4. In NZ, engage early. Work with Te Whatu Ora governance processes and keep an eye on the national research outcomes to guide adoption.
   

AI is not replacing radiologists, it is changing how we work and where we invest our time. The services that benefit most will be the ones that adopt carefully, measure outcomes, and keep patients at the centre of every decision.

How is AI showing up in your service today, and what would help you adopt it safely next quarter?

References

1. South Australian Medical Imaging. AI Chest X-ray Assistance Program Rollout. Government of South Australia, 2025.
2. I-MED Radiology Network. I-MED rolls out Annalise CXR nationally. I-MED Press Release, 2024.
3. Annalise.ai. Enterprise chest X-ray AI in Australia. Annalise AI Clinical Resources, 2024.
4. AHPRA. Regulation of Artificial Intelligence in Healthcare: Practitioner Responsibilities. Australian Health Practitioner Regulation Agency, 2024.
5. Therapeutic Goods Administration (TGA). Software as a Medical Device (SaMD): Guidance for Industry. Department of Health, Australia, 2023.
6. RANZCR. Position Statement: Artificial Intelligence in Radiology. Royal Australian and New Zealand College of Radiologists, 2023.
7. Te Whatu Ora. Artificial Intelligence in Health Advisory Work Programme. Government of New Zealand, 2025.
8. Office of the Prime Minister’s Chief Science Advisor (NZ). AI and Algorithms in Aotearoa: Opportunities and Challenges. 2024.
9. Massey University. Research Grant: AI Chest X-ray Evaluation in NZ Hospitals (2025–2027). Research Projects Database, 2025.