Healthcare systems in Australia are responding to complex and continually changing factors. With 28% of Australians living in rural and remote areas, together with a sizeable ageing population, many communities face a serious shortage of doctors and nurses. At the same time, disparities in health outcomes between Aboriginal and Torres Strait Islander (First Nations) people and non-Indigenous Australians are driving a 'health gap'. 

Many factors determine access to quality, affordable healthcare — not only in Australia, but around the world. According to the latest World Health Organisation (WHO) data, ''at least 50% of the world's population still lacks full coverage of essential health services''. Across Australia, the science, healthcare, and assessment industries are collectively looking at ways to tackle these issues and drive better healthcare outcomes for all. Last year, our national science agency — the Commonwealth Scientific and Industrial Research Organisation (CSIRO), launched research highlighting a new 'extraordinary era' of AI in healthcare.  As CSIRO's research director, Dr David Hansen, emphasised: ''The use of AI in healthcare is unique because the accuracy of models could mean the difference between life or death, or ongoing health or illness.'' 

According to Pearson's 2024 Skills Outlook, ''AI is set to transform clinician capabilities, improve diagnostics, and help deliver personalised treatments.'' To fully realise this potential, effective assessment and qualification will be essential for ensuring competency and fostering trust between practitioners and patients. So how can AI be used to measure and verify a wide range of skills required for safe and effective patient care?

Simulation-based assessments are commonly used to train healthcare professionals in their handling of real-world scenarios. Over the last decade, this method of assessment has evolved to evaluate technical skills across a wide variety of clinically related situations. Rather than a standalone tool, simulations are used alongside other tests of competence to ensure a healthcare professional is equipped to practise, and now — like many assessment tools — it's being reimagined by AI.

It will be exciting to see how simulation-based assessment — enhanced by AI — will allow healthcare professionals to demonstrate practical, real-world skills in much more dynamic and adaptive environments than ever before. While traditional real-world simulations have their benefits, they're expensive and can be limited in their use. Setting up a physical environment takes time and money — often actors need to be recruited and trained to play patient roles. And even then, there are some scenarios that realistically cannot be recreated — like rare emergencies or situations involving cultural nuances. AI can bridge that gap by generating more personalised, complex, or repeatable scenarios more quickly. Advancements with this technology will only broaden the use of simulation-based training and assessment in healthcare. 

Not long ago, NVIDIA released the world's first humanoid robot foundation model — along with advanced simulation frameworks. What's fascinating is that these robots can learn far more from simulated environments than from real-world experiences. And the same principle applies to human learning, too. When it comes to preparing healthcare professionals for high-stakes, fast-paced environments, AI-enhanced simulation might just be the most powerful training and assessment tool we have. Think for a moment about all the types of people you encounter in a hospital — professionals with varying degrees of experience — perhaps with different perspectives and from different cultures. With AI simulations, highly skilled, advanced medical professionals — as well as less experienced medical students, can rehearse different types of scenarios multiple times to mitigate risk or stress test new techniques or treatments.

The use of AI makes these examples of simulations more responsive to an individual's actions, offering a much richer assessment of their competency to handle complex challenges. By providing the professional with a more immersive experience in which to demonstrate their knowledge, skills, and abilities, AI-enhanced simulations are set to play a vital role in training and evaluating professionals in new job functions (as seen with the nursing example) as well as more complex medical procedures (such as those carried out by a surgeon).

While these simulations already exist in prototype form, the challenge within many healthcare settings will be how conveniently they can be developed, adjusted, and scaled — and at what cost? To be effective, AI-driven simulation-based assessments need to be targeted where they can have the most far-reaching impact. With rapid advances in technology, creating, and customising these simulations is likely to become much easier and more affordable in the near future.  

Simulation-based assessment, enhanced by AI, offers a powerful tool for training and evaluation — but like any other AI-assisted system, its output is only as good as the data it's built on. If that data contains any biases or misconceptions, then the AI will inevitably reflect and reinforce those same limitations. This is where the human touch (known as the "human-in-the-loop" model) will prove vital — particularly in clinical scenarios involving diverse patient populations. With health inequalities across different regions and different demographics, clinical simulation developers will need to be inclusive of all types of patients. And professional roles within this sector will still require a combination of technical expertise and essential human skills because they come with the highest stakes of all: people's lives. To enter 'an extraordinary era of AI', ensuring that any AI tools — including those used in simulation-based assessments — are fit-for-purpose, responsive to patient needs, and capable of tackling the most pressing health challenges will be fundamental.