Key facts
- The use of AI scribes by doctors in Australia has nearly doubled, from 22% to 40% between August 2024 and November 2025.
- The federal health department has identified concerns regarding the lack of oversight and potential privacy risks associated with AI scribes.
- Some AI scribe suppliers may be sending patient data outside Australia without explicit disclosure.
- The effectiveness and accuracy of AI scribes are subject to the same limitations as other large language models, impacting patient safety.
- Informed patient consent is crucial, requiring understanding of the technology's benefits and limitations.
- The Therapeutic Goods Administration is reviewing whether AI scribes should be classified as medical devices.
The Australian federal health department has issued a warning regarding the escalating use of AI scribe tools by doctors, citing significant privacy and oversight concerns. These AI tools, which record and summarize doctor-patient conversations for medical notes, have seen a surge in popularity, with usage among Australian general practitioners nearly doubling from 22% in August 2024 to 40% by November 2025, according to a poll by the Royal Australian College of General Practitioners (RACGP).
Concerns detailed in February 2026 Senate estimates briefing documents highlight that AI scribes "have little oversight" and raise risks for patient data security, particularly as some suppliers may be unaware their cloud platforms transmit data outside Australia. The department also noted that some tools are marketed with limited transparency, potentially outside regulatory oversight, and that advertised revenue increases for health professionals could have implications for the Medicare Benefits Scheme.
Further analysis from the department's AI advisory group in April indicated that while AI scribes can enhance clinician productivity and potentially reduce burnout, they share the quality and accuracy limitations of other large language models. This poses risks to patient safety, clinical accountability, and the integrity of national digital health data. A significant variation in obtaining informed patient consent for the use of these tools was also identified, with the department emphasizing that consent requires patients to understand the technology's benefits and limitations.
Consumer advocates echo these concerns, with the Consumer Health Forum stating that consumers and the government are asking similar questions about the adequacy of current safeguards. Reports from patients indicate instances where consent for AI scribe use was not adequately sought, with some providers refusing patients who did not consent. The oversight of AI scribes in Australia is fragmented, involving multiple regulators including the Therapeutic Goods Administration (TGA), the Australian Health Practitioner Regulation Agency (Ahpra), and the Office of the Australian Information Commissioner (OAIC). The TGA is currently reviewing digital scribes to determine their classification as medical devices, with a report expected in the coming months. The privacy commissioner is actively monitoring the situation, engaging with industry and civil society on consent protocols and privacy disclosures.