AI is the hot topic in health at the moment because it represents a foundational shift in how we operate. Australia is undergoing an AI revolution that will forever change how we work. Those choosing to harness the power of AI are giving many healthcare organisations a competitive edge. AI is automating routine tasks, and complex tasks are being solved more quickly. This means healthcare professionals can spend effort on high-value tasks and invest more time in patient care.

‍

Yet many of us are still using AI in a limited capacity and have barely wrapped our minds around how we can use it. Our industry must find ways to leverage AI to boost productivity and enhance patient care. Given that the technology and its potential uses in health settings are still in their early stages, AI best practices and risk mitigation are not yet established. Thus, there is a level of risk if it's abused or used incorrectly.

‍

While the rest of the world accelerates towards an AI-led future, Australia is in the infancy stages of AI development and implementation in healthcare settings. Countries like America and Asia are leading the way with experimental AI use and widespread adoption in healthcare and aged care settings. It's critical that we take the time to investigate which bottlenecks are preventing widespread adoption, or if it's even worthwhile.

‍

This article will explore some of the headline ways AI is being trialled and some cases adopted in different health settings. Now, let's take a look at AI being used in the aged care sector.

‍

AI in Aged Care

Australia’s well-documented ageing population, a trend seen worldwide, means the aged care sector is booming. It means we need to continue building both physical and digital infrastructure to meet service demand.

‍

Care is not always limited to physical care or capacity, but also to the wellbeing of our elderly. Loneliness has been a major problem experienced by aged care residents. Overseas robots and AI chatbots are being introduced to meet the social needs of aged care residents. Inspired by this uptake, Australian providers are introducing Abi and Daisy, in-home robots developed by the startup Andromeda. Abi and Daisy were designed to provide more frequent companionship while in their rooms. The robots are fluent in over 90 languages, recognise faces, remember past conversations, and participate in activities such as singing, dancing, and prayer.

‍

Going beyond companionship, the Federal Government is actively exploring safe and effective ways to use AI to enhance elder care. The Department of Health Disability and Aging is evaluating the role of AI in its five-year plan, focusing on research into its application in aged care settings. The Department is currently reviewing research, safety protocols and ethical guidelines, with plans for limited pilot programs involving up to 20 health professionals to test systems in controlled environments before any broader rollout. 

‍

AI healthcare settings

In Australian healthcare settings, AI is being used to improve communication and note-taking, with some doctors opting for digital scribes, multilingual support tech, and assessment tools to aid clinical diagnosis.

‍

These AI tools support the healthcare workforce by reducing the burden of administrative tasks, such as manually entering clinical notes, initial scanning and sorting of medical images, scheduling appointments, coding insurance claims, and responding to patient FAQs. When lower-value tasks get automated by AI, healthcare workers will have more time to spend with residents and patients. 

‍

Why are AI pilots being stalled in healthcare?

‍

With so many benefits AI can offer in healthcare and aged care, let's look at why the Australian healthcare industry remains hesitant to adopt this technology. While there are many benefits to unlock in the health sector, given its infancy, there are also many blockers to rapid change, for better or worse.

‍

We are seeing many healthcare companies, providers, manufacturers and government providers are keen to adopt AI, but many get stuck in the pilot stage. A recent study conducted by Appian surveyed 500 healthcare professionals: 60% of organisations were running small AI pilots, yet only 12% reported full deployment across multiple care or admin functions. The study highlights how fragmented data and disconnected systems made it difficult for AI to take hold.

‍

As a consequence, we are seeing a slow uptake of AI, and infrastructure is harder to establish and scale across the whole organisation. Appian also reports that 72% of healthcare workers regularly work with incomplete or inaccessible data, while 74% spend at least 1 day per week on manual administrative tasks.

‍

Healthcare data often gets trapped in silos, leaving healthcare workers stuck navigating fragmented systems. This creates a disconnect between the seamless experience patients have come to expect in the digital age, and the reality of the disjointed systems staff encounter daily. Every hour spent on admin takes time away from patient care.

‍

Overall, companies are willing to purchase AI tools but lack the time to integrate them. Addressing backend issues with AI could be key to alleviating the industry's capacity strain and removing low-value work, both of which pose a high operational risk to health services. Beyond fragmented systems, gaps in regulation and a lack of formal policy are slowing its uptake.

‍

The AI regulation gap

‍

AI has diverse applications, but the technology is still relatively new. Legal and regulatory frameworks are not robust enough to always protect users. Experts in this area, such as Alex Ruani, a doctoral researcher in health misinformation at University College London, are concerned about the dangers of this emerging tech.

‍

This year alone, we have seen the launch of ChatGPT Health in Australia, which is worrying many experts. Dr Ruani believes that ChatGPT Health is a transformative interface that can help people make sense of health information and test results, or receive dietary advice. But not a replacement for clinical advice. His primary concern is the risk of users being exposed to misinformation, as well as the potential for the AI to omit critical details such as side effects, contraindications, allergy warnings, and risks associated with specific supplements, foods, or diets.

‍

“There had been too many horrifying examples of ChatGPT leaving out key safety details like side effects, contraindications, allergy warnings, or risks around supplements, foods, diets, or certain practices”.

‍

ChatGPT is not a regulated model or regulated to give appropriate medical advice. There are no mandatory safety controls, no risk reporting requirements, no post-market surveillance, and no obligation to publish testing data.

‍

As a result, users may be misled into data sources containing misinformation or improper practices unrelated to their health conditions. Misinformation can directly put users at risk of further harm by causing stress and endangering them. At the same time, a clinician has the appropriate tools and knowledge to help patients. 

‍

A tale of two conditions and applications

As an industry, we are aware that the ways AI can mislead people about their health have fostered a certain wariness towards the technology. However, it has not halted curiosity about its potential applications. 

‍

A 2025 study, “Unravelling Public Preferences for the Use of Artificial Intelligence Mobile Health Applications in Australia”, examined public perceptions of AI diagnostics and virtual health assistants in relation to heart disease and depression. Over 1,100 Australians participated in a multiple-choice survey that evaluated five key characteristics: anonymised data sharing, human-AI interaction, AI accuracy, explanations of results, and funding sources.

‍

The survey results revealed that AI accuracy was most important to participants, followed by collaboration between human doctors and AI, rather than AI working alone. People preferred not to share anonymised data for depression, whilst no such preference emerged for heart disease. Explanations of AI results and funding sources were generally less influential in decision-making.

‍

Important differences emerged across groups: those who reported fearing AI were less likely to choose AI diagnostics or virtual health assistants for heart disease. Older adults (60+) showed lower acceptance of AI for both conditions, whilst younger adults (18-29) were more inclined to use virtual health assistants for heart disease.

‍

This research is very timely, with the federal government recently publishing the ‘Safe and Responsible Artificial Intelligence in Health Care Guideline’. Additionally, the Therapeutic Goods Administration has issued regulations for medical device software, including AI applications. Understanding user preferences for AI is crucial to ensuring that future AI technologies are implemented safely and fit for purpose. 

‍

Digital health systems are changing whether we like it or not.

Our government and related bodies are incrementally rolling out AI into our health systems. The CSIRO’s Australian e-Health Research Centre (AEHRC), with its scientists and engineers, has been working for over 20 years to make healthcare smarter and faster.

‍

AI and digitisation are already present in your doctor's office. Digital technology, such as ambient scribes, is now commonplace at GP offices and in health system platforms like My Health Record. Furthermore, the AEHRC has developed a new cardiovascular risk calculator that integrates directly with doctors' software to pull relevant data and record the results. The AEHRC is also trialling a digital care model with Queensland Health hospitals, replacing paper diaries with digital tools. These advancements are already happening and show we can optimise our healthcare system with technology if we take the time to work on it. 

‍

The AI Future 

AI has immense potential to extend beyond clinical settings, making healthcare more efficient and improving the wellbeing of patients and workers across the system. However, to make it work, we need to ensure careful consultation across the industry and robust regulations to keep patients and workers safe. 

‍

One of the biggest dangers of AI isn't even misinformation; it's the human element of care being diminished. Human care will always be at the centre of healthcare, and one thing…. machines cannot automate, replicate, or generate tender loving care.

‍

AI is the topic of this year's conference, and it's something we need to discuss as an industry. The AHSPO Conference provides us all with a great opportunity to discuss this. We look forward to having these discussions together.