DBrief: Summer Series — Executing AI in complex settings

Latest episode

In this episode, we spotlight two industries where the stakes couldn't be higher - healthcare and superannuation - and ask: how do organisations build trust, navigate regulation, and introduce AI responsibly in emotionally sensitive environments?

Our panel brings together three leaders implementing AI at scale:

• Greg Hill, Head of Data Strategy, Analytics and Insights, AustralianSuper, shares how an empathy-led, consultation-driven approach is helping uplift AI capability across investment, member, and enterprise teams - without overwhelming staff or compromising trust.
• Elizabeth Turner, Segment Director for Health & New Zealand, ISS Australia, reveals how her team tackled hospital food waste using computer vision, and why starting with the problem, not the technology, was the key to success.
• Rita Arrigo, National AI Centre, reflects on responsible AI, cultural fit when choosing vendors, and why visibility across AI projects can spark innovation and reduce duplicated effort.

From navigating regulated environments and cybersecurity, to building communities of practice and managing expectations around generative AI, this conversation explores the human, technical, and organisational dynamics needed to implement AI safely and meaningfully.

Key Takeaways

• Start with the problem, not the technology. Both organisations emphasised problem-definition as the foundation for successful AI adoption.
• Empathy-led change management works. Deep consultation, listening, and internal "consulting mindsets" help overcome fears and build genuine buy-in.
• Responsible AI requires cultural fit. Choosing vendors aligned with organisational values is as important as technical capability.
• AI can strengthen (not replace!) human judgment. Communities of practice, executive education, and bottom-up capability uplift ensure adoption sticks.
• Regulated environments can enable innovation. Standards like HL7 support safe data sharing while unlocking new clinical and operational insights.
• Integration is essential for long-term impact. Interfaces, workflows, and data governance must evolve to embed AI into core operations.
• Cybers AI is no longer bundled into the "cyber risk" bucket. Instead, leaders now see that AI is critical for defending against the rising volume and sophistication of cyber-attacks - providing detection, monitoring, and response capabilities that humans alone can't match.

Read Australian Industry Group's report Artificial Intelligence: Positive for companies, their people, and Australian Industry.

Note: This recording is from a live event and the audio quality may vary.

Contact the Industry Development & Policy team here.

Dive deeper into this topic by listening to our previous DBrief episode Summer Series: Getting Started - The AI Adoption Challenge for Industry.

Louise McGrath