Office of AI welcomed, but healthcare needs its own focus, says Enrico Coiera

10 minute read


Professor Enrico Coiera says Anthony Albanese's new Office of AI is a welcome leadership step, but warns healthcare needs dedicated AI policy, stronger sovereignty and sustained investment.


Professor Enrico Coiera, founder of the Australian Alliance for AI in Healthcare, has welcomed the prime minister’s decision to establish an Office of AI, describing it as an important leadership signal, albeit one that had arrived later than many in the sector had hoped. 

PM Anthony Albanese today announced the creation of a new Office of AI within the Department of the Prime Minister and Cabinet to coordinate a whole-of-government approach to artificial intelligence, alongside a national framework governing large AI data centres, copyright protections, and AI standards. 

Mr Albanese said Australia needed to move quickly to ensure AI served the national interest rather than allowing others to shape the country’s future. 

“It is not a question of if or when AI will transform our economy. We’re past that,” he said. 

“The question that matters, the choice that we have, is how – how we apply our enduring values of fairness and opportunity to make this technology work for us, for workers and communities, for our economy and our environment, for our creative industries and media.” 

Despite announcing the Office of AI, Mr Albanese made only a brief reference to healthcare – a motherhood statement about it “building new screening tools for cancer and disease”. 

Professor Coiera was untroubled by the lack of specifics about AI in healthcare, however. 

“The way to see today is to be happy that AI has now become a topic that is of core interest to the prime minister and therefore whole of government,” Professor Coiera told Health Services Daily

“That’s critical, and you’re not going to be able to imagine the transformations we hope for across healthcare without that level of commitment and engagement. 

“I’d personally say thank you for making this a national priority. That’s really important. It’s been a little bit late … it’s a difficult thing to engage in, and it’s moving so fast.” 

Professor Coiera said the PM’s announcement addressed a range of issues common across government, including security, privacy, and copyright, but healthcare would still require its own policy focus. 

“Healthcare does deserve special attention, and what I’m hoping is that this is the beginning of a ramping up of effort,” he said. 

“There is a lot to be done, there’s no question. If you look at this whole sector, the pace of change is substantial, and I don’t think moving at normal pace is going to get us there.” 

He pointed to the government’s recent work on digital scribes as evidence ministers recognised there was “a gap currently that they need to be filling”. 

Professor Coiera said healthcare faced challenges that went well beyond those confronting other sectors. 

“In healthcare, we have a parallel issue, which is use of our personal electronic health data to train models,” he said. 

“Companies might say that we are protecting your data, we’re treating your privacy front and centre. But what they’re not disclosing is how they might be using your data to train models and how much at risk you are down the track of being re-identified by a malicious actor.” 

He said healthcare also already had mature regulatory systems overseeing AI, meaning many of the sector’s challenges should continue to be led from within health rather than by the new Office of AI. 

“I would argue that healthcare, probably as a sector, has the most sophisticated regulations around AI because we are a high-risk sector and we recognise that if we’re not doing the right job, people will be harmed,” he said. 

“There’s a big piece to come around healthcare.” 

Professor Coiera said the alliance’s forthcoming AI roadmap – to be released next Wednesday – would argue Australia also needed to strengthen its AI sovereignty. 

“At the moment, the strategy for getting generative AI into healthcare is to adopt these large language models that are coming out of North America and to build our health systems around that,” he said. 

“That’s a very risky proposition, to base the operation of our health system on non-nationally controlled AI.” 

Instead, he said, Australia should ensure it retained control over AI capability and develop models that reflected Australia’s own health system, clinical practice, and population. 

“We need to have control over the capability,” he said. 

“In healthcare, that probably will mean having access to healthcare models which reflect the national characteristics, the national case mix, [and] the healthcare needs of Indigenous populations. 

“These are all different to models that come out of North America, for example, trained on North American data.” 

He said locally relevant AI models would also be essential because Australian clinical guidance often differed from overseas recommendations. 

“If you were to ask one of the large language models about should my child be vaccinated or not, if the large language model used an Australian guideline, it would give you a different answer to whether or not it used the American guideline,” he said. 

“For all those reasons, we need to be comfortable that we’re getting information that reflects the Australian position and need.” 

While Professor Coiera said the new Office of AI should provide whole-of-government leadership, he expected detailed healthcare AI policy to remain within the health portfolio, supported by agencies including the Therapeutic Goods Administration, the Australian Commission on Safety and Quality in Health Care, and the Australian Digital Health Agency. 

Looking ahead, he said the government’s biggest challenge would be backing its ambitions with sustained investment. 

“There needs to be substantial commitment of resource,” he said. 

“Even if we are in a world where we’re using North American models, there’s so much work that needs to be done to localise and to make them fit practice.” 

Office of AI  

Mr Albanese said the Office of AI would sit within his own department and bring together AI work already occurring across multiple portfolios. 

“It will work closely with the minister for industry and innovation, Tim Ayres, and the assistant minister for science, technology, and the digital economy, Andrew Charlton, to coordinate the design of our new Australian standards and to bring together the work that ministers right across the government are undertaking.” 

He said AI policy had until now largely been developed “issue by issue, sector by sector”, but argued the technology now required a coordinated national response. 

“Just as government developed coordinated approaches for other significant technologies, from civil aviation in the 1920s to genetics in the 1990s, we must do this with AI as well.” 

The prime minister outlined a broad range of portfolios already undertaking AI-related work, including energy, copyright, productivity, employment, education, communications, and national security. 

National AI standards 

The government will also develop a national regulatory framework for large AI data centres. 

Mr Albanese said the government had already announced “a set of expectations for large AI data centres” and would now bring those into “one regulatory framework, clear, consistent, and mandatory”. 

He said he would seek agreement from premiers and chief ministers at next month’s National Cabinet meeting. 

“We will aim to bring the legislation to parliament and carry it early next year,” he said. 

The framework will cover issues including planning, energy, water use, and community engagement. 

“It is not our goal to try and legislate for every possible eventuality or risk,” Mr Albanese said. 

“This is about having the flexibility to keep pace with change and get out in front of it.” 

He said the framework was intended to build public confidence while providing certainty for investors. 

“This is about building Australians’ confidence and trust in AI and our nation’s capacity to manage it, ensuring that our national interests and our national security are protected and providing the certainty for growth, for jobs, and for investment.” 

Other sector responses

Peter O’Halloran, the chief digital officer of the Australian Digital Health Agency, having any “guidance from the centre always makes life easier”.

“There are many challenges with AI that transcend one individual sector in the economy, whether it’s healthcare or finance or banking.

“Having a lot of those core challenges that we all face being dealt with once by somebody else, so that each sector and its own area can focus on their own specific unique challenges, is very beneficial.

“I think it will help healthcare greatly.”

Anja Nikolic, CEO of the Australasian Institute of Digital Health, also welcomed the Office of AI’s establishment.

“[It] will be critical to Australia’s ability to capitalise on the benefits of AI, and to mitigate the risks and issues it presents,” she said.

“The government has rightly recognised that Australia must have sovereign, nationally coordinated frameworks and standards overseen by a dedicated Office of AI.

“We are very supportive of the Prime Minister’s assurances about the management and regulation of data centres, which is now a worldwide concern. In Australia, the massive expansion of data centres in metropolitan and outer-metro areas requires national standards and coordination around location, water and power use, as well as social and health impact.”

Ms Nikolic urged the government to ensure healthcare-focused AI was central to the remit of the Office.

“AI is transforming healthcare in treatment, diagnostics, patient management and administration.  The potential benefits and risks are augmented in healthcare compared to many other industries – after all, we’re dealing with people’s health and wellbeing,” she said.

“We strongly advise the government to involve the Department of Health, Disability and Ageing, the Australian Digital Health Agency, Ahpra and the broader sector to properly consider the implications of AI for the healthcare system, workforce and consumers.

“We would like to see a coordinated, well-informed approach to ensuring the deployment and use of AI in Australian healthcare is safe, effective and equitable.”

Throughout the speech, Mr Albanese framed AI as an economic opportunity that should strengthen Australia’s sovereignty, productivity, and workforce rather than simply automate jobs. 

“We are serious about attracting frontier AI investment to Australia, because we want AI to support and create good jobs, not replace them,” he said. 

He concluded by arguing Australia should embrace AI while shaping its development around national values. 

“We should not treat AI as a threat to good jobs. We must use it as an instrument to help create them. 

“This is the responsibility, the challenge, and indeed the opportunity facing our country.” 

End of content

No more pages to load

Log In Register ×