The healthcare industry is no stranger to innovation. "Once in a generation" discoveries like penicillin, anaesthesia, and mRNA vaccines have had huge impacts on health outcomes. The smallpox vaccine alone is estimated to have saved over 200 million lives. Today, we are on the precipice of a medical revolution that could increase the frequency of these "once in a generation" scientific breakthroughs - the introduction of artificial intelligence. For an industry bogged down in administrative tasks, AI can serve as a trusty assistant to ensure highly skilled, talented medical professionals have more time to spend on patient care. Already medical research, diagnoses, operational procedures, clinical results, and patient care are beginning to leverage data and AI. For healthcare professionals, the automation of administrative tasks is already helping hospitals and primary care with workflow efficiency. In medical research, the ability to analyse massive data sets in previously unimaginable timeframes is supporting clinical studies. Recently, Stanford University (Nature, 2023) researchers used deep learning algorithms to analyse retinal scans, achieving a staggering 95 per cent accuracy in detecting diabetic retinopathy, a leading cause of blindness. The analysis of large datasets could also lead to early detection of health conditions, particularly in childhood neurological conditions which benefit from early intervention. Detection often relies on subtle cues that may go under the radar. In these cases, AI can help to pick up these early indicators. On a wider scale, AI also has the potential to completely change public health intelligence. In our own backyard, the South Australian government recently announced a $28 million program to implement AI applications across the public sector. Healthcare is a priority for the program, with evidence showing it can reduce costs, improve operational efficiency, and allow health workers to spend more time on clinical care. This is being replicated across the world, with hospitals, state governments and global groups such as the World Health Organisation (WHO) all expanding the use of AI to help capture, analyse, and forecast health trends. While Australia's AI ambitions are high, uncertainty remains over what shape federal legislation will take. There are calls from some quarters to adopt stricter, more prescriptive laws similar to the EU. Others argue in favour of leaning in and embracing the new AI frontier. Despite this, the federal government clearly sees AI's potential in healthcare with almost $30 million being invested for research into new ways AI could transform the sector. As organisations - including hospitals and healthcare providers - wait for policy clarity, they should put the foundations in place to hit the ground running once those questions are answered. For AI to reach its revolutionary potential, organisations need to have a robust data strategy in place. In healthcare, the benefits from the use of AI and machine learning are higher than in many other industries, but that means the stakes are higher, too. One of the biggest concerns in adopting AI is its potential for incorrect, inaccurate or potentially harmful predictions. We know the technology thrives on finding patterns in large volumes of data, but this is only possible when the data it consumes is clean, accurate and well-structured. That requires reliable and scalable infrastructure that makes data easy to use, simple to connect, and - above all - trusted. Yet in many healthcare environments, information is siloed across departments, facilities, and even countries. For AI to be implemented effectively, collapsing data silos into one location must be a priority. AI's predictive nature is restricted to the data it has access to. Biases, errors, or missing information in this data can lead to the AI model providing false information. READ MORE: The Centrelink recipients fighting for survival who you mightn't have heard of This is why public-facing AI models often make mistakes or hallucinate - the information they draw from is too vast and disorganised. In healthcare, such hallucinations could be life-threatening. AI tools in the healthcare industry must therefore be anchored by trusted, rigorously validated data sets to maximise accuracy and ensure the outputs are factual and dependable. Healthcare organisations, whether public or private health providers, health insurance, or medical research facilities, must all have a centralised view of the organisation's data to ensure any use of AI is gathering context and information from a complete and trusted data source. Some of Australia's largest healthcare firms, such as Telstra Health, are already expanding technology partnerships to ensure its digital ecosystem can support modernisation, without losing focus on customer care. Regardless of which route Australia's AI policy takes, AI is not going anywhere. In healthcare, more than any other industry, the benefits could be truly profound, particularly when it comes to medical research. By setting the data foundations in place, organisations can build on their AI success to date and extract the full potential the technology has to offer once policy questions are answered. Once all the pieces are in place, the next "once in a generation" discovery could be just a few clicks away.