AI is no longer just a future-of-radiology debate. In Australia and New Zealand, the real issue is whether teams are ready to use it safely, clearly and with the right clinical ownership.

Opportunities are still there. In many cases, more than ever. But what is changing is how radiologists are making decisions about those opportunities.

If you work in regional or rural radiology in Australia or New Zealand - or you are considering a move - you already know how stretched many services feel.

Most radiologists don’t worry about audits — until they have to. And when audits do happen, the outcomes often surprise people. Not because of concerns about clinical competence, but because of small gaps in documentation, assumptions about who is responsible, or professional plans that no longer reflect how someone is actually practising.

Every January, radiology enters a familiar rhythm: updated Medicare rules, revised item numbers, and fresh compliance requirements that shape how we work for the rest of the year. But 2026 brings more than a routine refresh. This year’s changes place greater emphasis on documentation quality, clinical justification and consistent reporting practices.

2025 has been a year of real movement for radiology across Australia and New Zealand. Some of that movement has brought welcome innovation. Some of it has brought pressure. And all of it points to a more complex, regulated and competitive landscape as we head into 2026.

Demand for imaging keeps rising, yet the workforce isn’t keeping pace. In rural areas, scan wait times are stretching. In metro hospitals, burnout is climbing. The “workforce crunch” has shifted from a prediction to a present reality.

ChatGPT-4, the latest version of Open AI’s massively popular dialogue-based artificial intelligence (AI) model, delivered a “low performance” when tasked with answering questions about echocardiography, according to a new research letter published in JACC: Cardiovascular Imaging.[1]

Ultrasound combined with MRI has the potential to enhance the differentiation of benign and malignant breast nonmass-like lesions, according to the results of a study published in Ultrasound in Medicine & Biology. [1]

The GPT-4 artificial intelligence model, the backbone of popular chatbot ChatGPT, showed it can generate radiology reports that are comparable in quality to radiologists’ reports at a more efficient rate.