As AI gradually becomes part of everyday operations in the Australian aged care sector, we’re seeing growing anxiety that the more efficiencies we try to drive with AI in aged care homes, the less human they may feel.
Of course, aged care is the last place efficiency should overtake connection, where sometimes all residents need is someone to sit with them, listen to them and know their name.
But the idea that efficiency and humanity are in competition doesn’t hold up. In fact, in most aged care homes today, the biggest threat to human connection is the sheer volume of admin consuming a carer’s attention. This can become a quality issue, because the more time spent working on documentation, the fewer opportunities carers have to notice small changes in the health of residents early.
In this case, using AI to create efficiencies could actually be the solution to bring back the human touch by reducing that load for carers, if it’s done in the right way.
That’s why, the real questions for the sector undertaking large-scale digital transformation should focus on ‘how AI can give time back to workers for caring?’, ‘what AI should be trusted to do in aged care homes?’, and ‘what must always remain in human hands?’
Where the human touch actually lives
Answering these questions starts with distinguishing between the act of processing information versus exercising judgement on it.
Clinical judgement in aged care is built on the unique knowledge and understanding of a resident, which a carer acquires over time spent with that person. From changes in behaviour and mealtime habits to what a ‘good’ or ‘bad’ day looks like for that resident, carers become attuned to this, and no algorithm will ever compare.
Therefore, when it comes to AI implementation, it means AI should support the people delivering care but not replace their judgement or their authority to act.
AI helps most by empowering carers to apply their clinical judgement, quickly reading, sorting and surfacing the exact information needed for better decision-making. This allows carers to spend more of their shift observing residents, speaking with families and catching subtle health changes early, before they escalate.
This is a meaningful distinction the sector needs to get right from the start. By keeping AI informative and organisational, we prevent scope creep from ‘informing decisions’ to ‘making decisions’. This ensures clinical interpretation and decisions remain entirely in the hands of the people best placed to protect the trust and relationships at the heart of aged care.
Starting where the labour is heaviest
A natural starting point is looking at where the administrative burden is heaviest but does not involve clinical decision-making, and where AI can help to ease this burden on clinicians.
Handovers are a great example. Every shift at a residential aged care home begins with the handover, providing a clinical picture for every resident across the past 24 hours. An essential task, but an incredibly time-consuming one. At any given Regis Aged Care home, this could mean working through over 200 pages of clinical notes for up to three hours before a shift has even properly begun.
None of that time is spent making clinical decisions; it is spent finding the information needed to make them. When a carer is buried in paperwork, they may miss a change in tone, appetite or behaviour – the earliest indicators a resident needs attention. Every minute taken to read through handover notes is a minute less spent with residents, observing them and making vital care decisions.
This makes handovers a legitimate, rather than risky, task for AI to support, provided it can do so to the highest standard.
Setting the bar from the outset
When Regis and Cognizant partnered to roll out AI nationally across Regis Aged Care homes, we put three conditions that had to be met to make the rollout a success. The AI had to surface every resident and clinical issue with zero tolerance for hallucination; it had to sit safely within existing systems frontline staff already relied on; and it had to be intuitive enough that time-poor staff would actually choose to use it, rather than work around it.
The most effective way to uphold these standards, was to benchmark AI-generated summaries against a gold standard set by a senior panel of clinicians, testing against categories, such as agitation, medication and infection, and comparing performance to a registered nurse baseline on an ongoing basis.
Additionally, we knew that trust wasn’t a given simply because the tools were in place; it had to be built over time through transparent testing and honest change management with the people expected to rely on it day to day.
With these standards in place, Regis was able to reduce handover reviews from up to three hours to less than one hour, with no hallucinations recorded through the trial process and note-review accuracy matching or exceeding the registered nurse baseline across most clinical categories. By meeting these standards and building that trust, adoption then followed and most staff are now frequent users, engaging with the tool multiple times a week.
Building AI around the organisation
This example helps us demonstrate how AI needs to be built for a setting like aged care. Generic models cannot understand the nuances of a night shift, the differences between routine and urgent documentation, or the regulatory obligations that shape how care must be recorded. Without that context, AI in aged care is a liability.
This underpins what Cognizant would describe as being an AI builder, considering how deliberately the AI model is constructed around an organisation’s real workflows, responsibilities and risk tolerance from the start. Acquiring an AI platform and just hoping it fits will never allow an organisation to scale that tool safely and effectively across a large network and risks leaving that tool firmly in the pilot stage.
Letting AI do the admin, so people can do the caring
As more aged care and healthcare providers look to AI to manage rising demand and workforce pressure, the industry must focus on solutions that give residents more high-quality attention, while restoring the time and headspace carers need to make critical care decisions.
Handled deliberately, AI and human connection in aged care won’t be in competition. Instead, they will reinforce each other by letting AI handle the admin and giving carers better information, faster, so they can focus more of their time on delivering safer and more personal care.
Add Comment