Australia raises privacy concerns as doctors’ use of AI medical scribes surges
Australia raises privacy concerns as doctors’ use of AI medical scribes surges
Publish Date: 2026-07-05 06:43:00
Source Domain: ilkha.com
Australia’s federal health authorities have raised concerns over patient privacy, data security and clinical accuracy as the use of artificial intelligence-powered medical scribes expands rapidly across the country’s healthcare system.
AI scribes record conversations between doctors and patients, automatically transcribe consultations and generate clinical notes, reducing the administrative workload for healthcare professionals.
According to a survey by the Royal Australian College of General Practitioners (RACGP), the proportion of doctors using AI scribes increased from 22% in August 2024 to 40% by November 2025, highlighting the technology’s rapid adoption.
Documents prepared for the Senate and obtained by Guardian Australia under freedom of information laws show the Department of Health warned in February that the growing use of AI scribes has outpaced regulatory oversight.
The department said some providers market their products outside Australia’s medical device regulations while offering limited transparency about how patient information is processed and stored. Officials also warned that sensitive health data could be transferred to cloud servers located outside Australia, creating potential privacy and security risks.
Authorities further expressed concern that some companies promote AI scribes by claiming they can increase doctors’ income by up to 30% without extending working hours, raising questions about possible additional costs for the country’s Medicare system.
A separate assessment by the department’s Artificial Intelligence Advisory Group acknowledged that AI scribes could improve efficiency and reduce physician burnout. However, it warned that the limitations of large language models could pose risks to patient safety, clinical accuracy and professional accountability.
The report also found significant differences in how healthcare providers obtain patients’ consent before using AI-assisted…
