The final stop of our five-city series brought together four leading voices from across the care sector to unpack what AI really means for aged care the wins, the challenges, and the white space ahead. Here's what we learned.
Our expert speakers:
- Dr Paul Nicolarakis, CEO, Argentic Labs
- Petrina Greenwood, General Manager Innovation, Research and Partnerships, BaptistCare
- Alissa Brown, Chief Performance Officer, Apollo Care
- Jowe Esguerra, Digital Innovation Lead, Uniting NSW/ACT
1. We're Earlier Than We Think And That's Okay
The AI revolution in care is in its infancy. When global AI adoption is mapped against internet adoption curves, we're sitting somewhere around 2005 in internet years pre-Uber, barely Facebook. The implication? We shouldn't feel bad for not having it all figured out, but we also can't afford to be complacent.
"We are really early in the journey... if we want to get a sense of where we are heading, look at the sectors that have actually found their happy place with AI."
Dr Paul Nicolarakis, Argentic Labs
That framing should be genuinely reassuring for care sector leaders who feel behind. The organisations that have found their footing with AI software engineering being the most dramatic example, where tasks that once took six months now take minutes got there through experimentation, iteration, and a willingness to sit with uncertainty. The care sector doesn't need to have all the answers yet. It needs to stay curious and keep moving.
2. Strategy First. Technology Second. Always.
The temptation to reach for AI as a solution before defining the problem is real and dangerous. The organisations seeing the most meaningful results are those that started with a clear strategic challenge and then asked whether AI could help solve it, not the other way around.
"AI can't be the end game really you need to look at where your problems are and then look at implementing it. The technology should follow strategy, not the other way around."
Alissa Brown, Apollo Care
This is especially relevant in aged care, where the two biggest systemic pressures financial sustainability and workforce supply are well understood. The question isn't whether AI can help with those challenges. It's about being precise enough to identify exactly where in your operations it will have the most leverage, and building your AI investment case around that specificity rather than chasing the technology itself.
3. Garbage In, Garbage Out Data Hygiene Is Non-Negotiable
Before any AI tool can deliver value, the data feeding it needs to be clean, structured, and trustworthy. Organisations that skip this step are simply amplifying existing problems, not solving them. Getting data foundations right isn't a precursor to the AI journey; it is the first step of the AI journey.
"Good use of AI wouldn't be meaningful without proper information architecture. Garbage in, garbage out models are trained on data."
Jowe Esguerra, Uniting NSW/ACT
There's a harder truth embedded in this too. AI models are trained on human-generated data, which means they inherit human bias. If your data reflects historical inequities, inconsistencies, or gaps in care documentation, the AI will reflect those back at you with confidence. Investing in data governance and hygiene isn't a back-office IT concern; it's a care quality concern.
4. Fear of AI Is Usually Bigger Than the Reality
Across multiple organisations, panellists reported that the anticipated fear and resistance from frontline staff particularly around privacy and job security consistently turned out to be smaller than expected once people actually started using the tools. The narrative we build around AI matters enormously.
"Overwhelmingly, the perception of fear that we think our staff are going to have in my experience, the perception is greater than what I have actually experienced."
Petrina Greenwood, BaptistCare
That doesn't mean fear should be dismissed or minimised. It means the most effective response to it is exposure, not reassurance. When staff can see the tool working, understand what it can and can't access, and experience it making their day easier rather than threatening their role, trust follows. The organisations making the most progress are investing as much in the human rollout as in the technology itself.
5. Real Results Are Already Here The Numbers Speak
AI-driven decision support is already delivering measurable operational outcomes in aged care. Apollo Care's AI-powered Operations Insights Suite helped site managers understand workforce patterns, forecast costs, and make smarter rostering decisions with striking results.
"Across the board, our financial sustainability has improved by 28%. We've taken our agency staff use from almost 30% to 4%."
Alissa Brown, Apollo Care
These aren't pilot numbers, they're organisation-wide outcomes achieved by giving site managers better visibility over their own data and the analytical support to act on it. The insight that drove the agency reduction wasn't radical: managers could finally see exactly when and where agency staff were being deployed, understand the cost impact in real time, and make smarter scheduling decisions as a result. The AI didn't replace their judgment. It made their judgment better.
6. Frontline Staff Need to Be Brought Along Not Just Told
Successful AI implementation isn't a technology project; it's a people project. Frontline workers need to understand what's in it for them, be involved in the design process, and have their fears genuinely heard and addressed. Top-down mandates without human-centred change management are a recipe for low adoption.
"What is the core problem that we're trying to solve? And if you have the right answers to that without getting lost in the hype of AI make sure that through that process you don't forget about the human, the people who will be adopting it."
Jowe Esguerra, Uniting NSW/ACT
Uniting's experience with their digital assistant 'Buddy.' is instructive here. The team didn't just build a tool and roll it out they worked with language subject matter experts, studied their workforce demographics, and built the interface to accommodate the linguistic reality of their frontline. When workers feel seen in the design of a tool, adoption follows. When they don't, even good technology sits unused.
7. The Human Must Stay in the Loop
Across every use case discussed from workforce analytics to clinical note-taking the panel was unanimous: AI supports human decision-making; it does not replace it. The moment an organisation removes the human check from an AI output, risk escalates significantly. Test, verify, and judge.
"You can't use AI fully autonomously. You need the person, the human, in the loop. It can suggest and prompt and alert, but it's still the person making the decision."
Alissa Brown, Apollo Care
This principle becomes even more critical as AI models become more capable and more convincing. Hallucination where a model produces confident but entirely fabricated outputs remains a live risk, and the antidote isn't just better models. It's building cultures where staff know to interrogate AI outputs rather than accept them, and where workflows are designed with verification steps built in as standard practice.
8. Voice AI Is Unlocking a New Frontier for Care
Text-based interfaces create barriers particularly for a diverse, multilingual aged care workforce and for older residents who don't type. Voice AI, now nearly indistinguishable from human speech in quality, is opening up a genuinely new category of care interaction: more natural, more accessible, and more human.
"As a species, we've only written for about 5,000 years. We've spoken for about 250,000 years. When you're asking someone whose first language is not English to write, that is a much harder ask than just asking them to speak."
Dr Paul Nicolarakis, Argentic Labs
The practical implications for aged care are significant. Voice interfaces lower the documentation burden for workers whose written English is a second or third language. They open up direct engagement channels for residents who find screens and keyboards alienating. And they make it possible to capture the texture and tone of a care interaction not just a text summary of it which may ultimately tell us far more about the quality of care being delivered.
9. The "White Space" Is Where the Real Opportunity Lives
Beyond efficiency gains, the most exciting AI applications are the ones that were simply impossible before things no one even tried because the cost and effort were prohibitive. Building a rich life biography for every aged care resident. Calling hundreds of clients during a heatwave. Giving consumers their own record of a clinical consultation. This is the white space, and it's vast.
"These white spaces are where the real excitement's going to be, the real sizzle... a much more curious and complex and human-driven prospect that could never be done in an old world where it was just manual work."
Dr Paul Nicolarakis, Argentic Labs
The shift this represents is profound. For decades, care organisations have been optimising what already exists making existing processes faster, cheaper, or more consistent. White space thinking asks a different question: what would we do for residents and clients if cost and effort were no longer the constraint? The answers to that question are where AI's most meaningful contribution to care quality will ultimately be found.
10. Impact Has to Be Demonstrated Boards Are Watching
AI investment in care is no longer just a pilot conversation. Boards and CFOs are beginning to ask hard questions about return on investment, and organisations that can't demonstrate tangible impact on residents, staff, and financial sustainability will struggle to justify continued spend. Measuring outcomes isn't optional it's the price of continued innovation.
"At the end of the day, if we don't actually make an impact on the lives for the better of our staff and residents, why do it? The time is coming where they're going to say: this is how much we've spent on this what has been the impact?"
Petrina Greenwood, BaptistCare
This is the accountability moment the sector is moving toward, and it's a healthy one. The organisations best positioned to meet it are those that defined success metrics before they implemented not after. If your AI initiative doesn't have a clear theory of change that connects the technology to a resident, workforce, or financial outcome, that's worth addressing now, before your board asks the question for you.
Thank you to our Sydney panellists and to everyone who joined us across all five events in the Building Workforce Capability with AI series. The conversations, questions, and energy in the room at every stop reminded us why this work matters.
