The words 'artificial intelligence' do not appear anywhere in the strategy. Not once.
I spent some time reading the newly released 2026-2031 National Health Reform Strategy, endorsed by Australia’s health ministers earlier this month.
There is a lot to like.
The strategy correctly identifies many of the structural problems clinicians and patients experience every day: fragmented care, increasing complexity, workforce pressure and a health system that still struggles to move patients and information seamlessly between organisations.
More importantly, several of the national priorities are exactly where I think health reform needs to focus: improving patient flow and integration between care systems, designing and expanding better models of care, delivering continuity across the care journey, modernising and connecting our systems through data and digital health, and measuring whole of system performance.
That is an important shift in emphasis. Digital health is not positioned simply as an IT program. It is being considered as part of how we improve the performance, efficiency and connectivity of the health system.
From my perspective as a paediatrician and the founder of Consultmed, where we are building a national digital referral and advice-first network, this is particularly encouraging.
We have long believed that one of the biggest opportunities in Australian healthcare is not simply digitising individual hospitals or practices. It is connecting the journeys between them.
A GP referring a patient to a public hospital. A hospital redirecting that patient to a more appropriate service. A specialist providing advice without requiring a face-to-face appointment. A regional service collaborating with a tertiary centre. A discharge summary or specialist letter reaching the clinicians who need it. Patients being able to follow what is happening rather than disappearing into an administrative black box.
That connective layer is where enormous amounts of waste, duplication, delay and frustration still exist.
The strategy also recognises that better models of care need to be identified, evaluated and then scaled, rather than every jurisdiction repeatedly designing its own pilots.
Please. healthcare leaders, no more pilots of proven technology.
That could become particularly important for Advice & Guidance, virtual specialist care, community-based care and other models that give patients access to the right expertise without automatically adding another person to an outpatient waiting list.
There are, however, some striking omissions.
Where is artificial intelligence?
The words artificial intelligence do not appear anywhere in the strategy. Not once.
For a national health reform strategy extending to 2031, that is difficult to ignore. AI will undoubtedly reshape almost every part of this reform agenda.
The digital front door to healthcare is already changing. Increasingly, patients are turning first to tools such as ChatGPT, Claude and Gemini to understand symptoms, interpret information and decide what to do next. That has profound implications for access, navigation, health literacy and how people enter the formal health system.
Inside health services, AI is already changing how information is documented, summarised and interpreted, and is beginning to influence triage, diagnostic support, imaging, pathology, prescribing, correspondence, coding, scheduling and administration.
At a system level, it will increasingly affect demand forecasting, capacity allocation, measurement of outcomes and how limited workforce and funding are deployed. It may also fundamentally change models of care by making high-quality information, decision support and specialist guidance available much earlier in the patient journey.
There are profound workforce implications as well. I wrote recently in Health Services Daily about the risk of never skilling a generation of clinicians. AI may give clinicians enormous amounts of time back, but we also need to consider what happens when students and trainees use AI to perform tasks during the very period in which they are meant to be developing the underlying reasoning, judgement and expertise themselves.
An AI copilot is extraordinarily useful, but only if we have first taught people how to fly.
At a national level, many of the policy settings remain unresolved. Australia needs clear positions on data sovereignty, privacy, AI regulation, national guardrails, clinical governance, model assurance and procurement standards, as well as greater clarity about clinical responsibility and liability when AI contributes to a decision and where human oversight remains necessary.
For a strategy extending to 2031, the absence of AI is therefore particularly striking.
Related
We talk about connection, but interoperability still depends on execution
Interestingly, the word interoperability also does not appear in the strategy.
The intent is certainly there. The strategy talks about connected systems, Healthcare Identifiers, discharge information, authorised data sharing and a future National Health Data System.
There is also important work already underway through the Australian Digital Health Agency, particularly Health Connect Australia and Provider Connect Australia. If successful, these initiatives could become a pivotal part of the national digital health backbone.
But connectivity does not happen simply because the infrastructure exists. It requires APIs, standards, identifiers, directories and, importantly, vendors implementing them within the systems clinicians use every day.
That is why technology vendors need to be involved early in the design and implementation of this infrastructure, rather than treated as an afterthought. Ours are ultimately the organisations that have to translate national standards and infrastructure into functioning clinical workflows.
Australia increasingly has the foundations for interoperability. The next challenge is turning them into connected experiences that clinicians and patients appreciate.
Digital health is positioned where it belongs: under ‘value’
One thing I particularly like about the strategy is where digital health has been positioned.
Value is one of the five overarching outcome domains, defined around innovation, efficiency, effectiveness and value for public investment. Within that domain sits the national priority to modernise and connect our systems through data and digital health.
I think that positioning is really important. I have taught value-based healthcare in clinical practice at UNSW and ANU across medical and postgraduate programs. One of the core principles I try to emphasise is that value is not simply about reducing cost. It is about improving the outcomes that matter to patients relative to the resources required to achieve them. That is also the right lens through which to assess digital health.
An AI scribe is not valuable simply because it saves keystrokes. Its value lies in whether it gives clinicians meaningful time back, improves the quality of documentation and allows more attention to be directed towards patients.
Virtual care is not valuable simply because a consultation occurs online. It should improve access, convenience, continuity or outcomes without compromising quality.
Triage tools should help patients reach the right care sooner. Referral systems should reduce delays, duplication and administrative friction. Decision support tools should improve the quality and consistency of care.
The goal is not more digital health. The goal is better healthcare, with digital tools helping us deliver greater value for patients and the health system.
Overall, I think the strategy sets a strong direction. Patient flow, better models of care, continuity, connected systems, whole of system performance and value are exactly the conversations we should be having.
And by 2031, I suspect AI will be a fairly important part of that story.
Vikram Palit is a paediatric respiratory physician founder and CEO of Consultmed, and associate professor of paediatrics and child health at Western Sydney University.
This article was first published on Professor Palit’s LinkedIn feed. Read the original here.