Artificial intelligence, disclosure and consent: a comparative review of Australian and European Union legal frameworks for medical imaging.
Authors
Affiliations (2)
Affiliations (2)
- Teaching Innovation in Radiography Group, Charles Sturt University, Wagga Wagga, NSW, Australia; School of Dentistry and Medical Sciences, Charles Sturt University, Wagga Wagga, NSW, Australia; College of Health, VinUniversity, Ha Noi, Viet Nam; School of Medical and Health Sciences, Tung Wah College, Hong Kong. Electronic address: [email protected].
- Centre of Law and Justice, Charles Sturt University, Canberra, ACT, Australia.
Abstract
To conduct a comparative narrative review of Australian and European approaches to transparency, disclosure, and informed consent in AI-assisted medical imaging, with attention to patient autonomy, professional responsibilities, and regulatory oversight. AI systems are increasingly integrated into medical imaging workflows and offer significant benefits in diagnostic speed, efficiency, and accuracy. However, the reviewed literature highlights substantial concerns relating to transparency, algorithmic bias, automation bias, informed consent, and accountability. Compared with Australia, the European Union has more explicit statutory requirements for transparency, information provided to AI deployers, human oversight, and the governance of high-risk AI systems. Australian regulation remains distributed across privacy law, medical device regulation, common law, state consent legislation, and professional guidance. Current regulatory approaches to AI-assisted medical imaging remain inconsistent internationally. European legislation currently represents a stronger and more comprehensive model for disclosure and transparency obligations compared to Australia's developing framework. Ongoing legal reform, interdisciplinary collaboration, and responsive governance structures will be necessary to address the rapid advancement and increasing integration of AI systems within healthcare. Disclosure should be proportionate to the role, autonomy, risk, and clinical significance of the AI system. Where AI materially affects an examination or diagnostic decision, patients should receive clear information about its involvement, the extent of human oversight, material risks and limitations, how their data may be used, and any reasonably available alternatives. Radiographers and radiologists require appropriate AI literacy and clear local governance to support consistent communication, professional judgement, and accountability.