top of page

Auckland University of Technology - AI in Health Research Hui 2025

Updated: 8 minutes ago

In April 2025, Auckland University of Technology (AUT) hosted a session exploring AI in health — covering research, policy developments, ethical considerations, and opportunities for collaboration.


Assoc. Prof Rosie Dobson, Dr. Kevin Ross, and Assoc. Prof. Mangor Pedersen
Assoc. Prof Rosie Dobson, Dr. Kevin Ross, and Assoc. Prof. Mangor Pedersen

Speakers included:

  • Robyn Whittaker, who gave an update on the state of AI at HealthNZ and

    their research priorities.

  • Kevin Ross from Deloitte, who provided an industry-focused perspective, outlining current initiatives, directions, and opportunities for AI across New Zealand.

  • Mangor Pedersen from AUT, who spoke on the Australian Epilepsy Project and responsible AI use.

  • Catherine Shi, Head of the new Department of Data Science and AI at AUT, who shared her research journey, including the development of computational methods to analyse multimodal brain imaging data.

  • Rosie Dobson, who presented on the Artificial Intelligence in Health Research Network which she chairs.

 

A recurring theme was the importance of collaboration among stakeholders — including clinicians, industry partners, academics, and policymakers — to navigate the transformation AI can bring. Shared solutions are needed for

challenges such as data sharing, model adaptation, and equitable access.

 

The presentations were followed by a panel discussion. One particularly interesting point was the risk of perpetuating biases in datasets used for AI models.

 

Kevin Ross offered the following insight - “The first thing you start to do with AI is to automate what you've already done, and there are dangers in terms of perpetuating historical inequities. But if you take a slight step back from that, what you're actually doing is identifying and quantifying inequities — and that's a good place to start… So, through this technology, we have an opportunity to recognise what happens today and build a layer of AI that really helps with the problem.”

 

If you're interested, a full recording is available on the AUT website.

bottom of page