Catholic Health Australia has urged practitioners to exercise careful management in the use of Artificial Intelligence technologies that it says can offer significant benefits but can also carry novel risks that could cause significant harm.
As is the case in many industries and sectors, AI is transforming how healthcare is delivered across Australia. Some of the many applications currently in use include scribes in emergency departments, AI-powered imaging in cancer diagnostics, predictive algorithms in surgical planning, and falls-prevention monitoring in residential aged care.
“AI has enormous potential to improve diagnostic accuracy, reduce administrative burden on clinicians, and expand access to care for underserved communities,” explained CHA Director of Mission, Brigid Meney.
“Those benefits,” she continued, “are not guaranteed unless health and aged care organisations put robust governance in place before deploying these tools.”
Ms Meney explained some of the risks associated with the use of AI include, “undermining the clinical encounter between patient and professional, and the propagation of misinformation and errors or ‘hallucinations’ which can have severe consequences, ranging from misdiagnosis to delayed treatment, adverse drug reactions, and inappropriate clinical decisions.”
In a recently released position paper on the issue, CHA has also warned AI algorithms can replicate and entrench biases in training data, producing worse outcomes for Aboriginal and Torres Strait Islander peoples, older Australians, people from culturally and linguistically diverse communities, and people in regional and remote areas.
“The populations most likely to be harmed by poorly governed AI are precisely those our mission commits us to serve,” Ms Meney said.
“As Pope Leo XIV said in his encyclical Magnifica Humanitas, when an AI system presents itself as neutral and objective, it can mask the exclusion of the most vulnerable behind a veneer of neutrality, making injustice harder to see and to challenge.”
CHA said the Catholic sector, which shares a common mission and ethical framework, could work together to manage the risks of AI including appointing a dedicated person responsible for AI adoption, conducting formal equity impact assessments before deploying any new tool, and investing in AI literacy for clinical staff.
“No AI tool should be deployed unless its impact on the most marginalised people using that service has been formally assessed,” Ms Meney suggested.
