AI Scribes Are Listening In on Your Doctor Visits: New Privacy Warning and What to Do
You sit down in the exam room and start describing your symptoms. The doctor listens, asks questions, and types. But more and more, that typing isn’t just the doctor’s fingers on a keyboard—it’s an artificial intelligence system transcribing your entire conversation and automatically generating medical notes.
AI scribes are now common in clinics and hospitals. They promise to save doctors time and let them focus on you rather than on paperwork. But a recent warning from the Australian government has raised serious questions about what happens to that recording of your private medical conversation. If you live in any country where these tools are being adopted, the implications are worth understanding.
What happened
In July 2026, the Australian government issued a formal privacy warning about the use of AI scribes in healthcare. The warning, reported by The Guardian, highlighted that many patients are not adequately informed when these tools are used, and their sensitive health data may be stored on third-party cloud servers, shared with AI vendors, or used to train future models without explicit consent.
The Australian Office of the Australian Information Commissioner (OAIC) reportedly expressed concern that doctors may not be fully complying with privacy laws when using AI transcription services. The core issue: patients typically have no idea an AI is listening, and their consent is often not sought in a meaningful way.
Why it matters globally
Medical conversations contain some of the most sensitive personal information you have: diagnoses, medications, mental health history, family medical data, even lifestyle habits. In the wrong hands, that information could lead to discrimination, insurance problems, or embarrassment.
AI scribes are not just passive recorders. They process speech through cloud-based services that may be operated by companies like Microsoft (Nuance DAX Copilot), Amazon (AWS HealthScribe), or startups. These companies’ privacy policies and data retention practices vary widely. Some may keep transcripts to improve their AI models, and unless contracts explicitly forbid it, your data could end up in training sets.
The Australian warning is a signal that even governments are catching up to the reality: the convenience of AI scribes is outpacing regulatory safeguards. Other countries, including the United States and United Kingdom, are also seeing rapid adoption, but patient protections often lag behind.
What you can do
You don’t have to accept AI transcription silently. Here are practical steps to protect your privacy during a doctor visit:
Ask directly. Before your appointment begins, ask: “Are you using an AI scribe or any tool that records or transcribes our conversation?” Many doctors are upfront if asked. If unsure, request clarification.
Read consent forms slowly. Some clinics include AI scribe use in the general consent form you sign at check-in. Look for phrases like “audio recording,” “transcription service,” or “third-party data processing.” If you see these and don’t agree, you have the right to ask to opt out.
Request an opt-out. In most jurisdictions, you can refuse to be recorded. The doctor should still be able to take notes manually or use a non-AI system. If they say it’s not possible, consider whether that practice respects your privacy preferences.
Ask about data storage and deletion. Even if you agree to AI transcription, you can ask: “How long is the transcript kept? Is it stored on a cloud server? Will it be used for training?” Doctors may not know all the details, but your questions signal that you expect transparency.
Consider using a privacy-focused doctor. Some clinics now advertise that they do not use AI scribes or that they use only locally processed, on-device tools. If this is important to you, research your options beforehand.
Future outlook
Regulations are likely to catch up. The Australian warning may prompt other governments to issue similar guidance or even mandate patient consent for any AI-assisted transcription. In the meantime, the burden remains on patients to stay informed and ask questions.
AI scribes are not inherently bad—they can reduce burnout and improve note accuracy. But like any technology that handles sensitive data, they deserve scrutiny. Your medical privacy is not a convenience trade-off that should be made without your knowledge.
Sources
- The Guardian, “Doctors’ soaring use of AI scribes prompts Australian government warning over privacy,” July 4, 2026. (Article referenced in the evaluation research.)