AHPRA's AI guidance does not create new rules. It explains how health practitioners' existing professional obligations under their Code of Conduct apply when they use AI tools such as AI scribes. Practitioners stay accountable for every output, must understand the tool's limits, should tell patients when AI is used, and must apply human judgement before anything enters the record.
Key takeaways
- The guidance from AHPRA and the National Boards, "Meeting your professional obligations when using Artificial Intelligence in healthcare", confirms that existing obligations already cover AI.
- A note drafted by an AI scribe is the practitioner's note. "The software generated it" is not a defence.
- Patients should be told when AI is used in their care and offered a workable alternative if they decline.
- Before deployment, confirm where patient data is processed and stored, and whether it is used to train vendor models.
- Pilot with close review, document your consent process and review rule, then scale with ongoing audit.
What does AHPRA's AI guidance say?
AI scribes that draft consultation notes, triage assistants, coding helpers and patient-messaging tools are already in general practices, allied health clinics, dental groups and specialist rooms across Australia. The technology moved first, and professional expectations are catching up.
The Australian Health Practitioner Regulation Agency (AHPRA) and the National Boards have published guidance titled "Meeting your professional obligations when using Artificial Intelligence in healthcare." It is short and deliberately avoids becoming a new rulebook. It explains how the obligations practitioners already hold under their profession's Code of Conduct apply when AI enters practice. For practice owners, clinical governance leads, practice managers and registered practitioners in health groups, it is the frame the regulator will use if a question is raised.
The guidance is no reason to avoid AI. Used well, an AI scribe can reduce the after-hours documentation load that contributes to clinician burnout. Accountability simply stays where it was.
AHPRA's four principles for AI use
- Accountability: you remain responsible for your practice and must apply human judgement to any output of an AI tool.
- Understanding: understand enough about the AI you use, including how it is trained and tested, its intended use and its limitations.
- Transparency: be open with patients about your use of AI, and address any concerns they raise.
- Overall responsibility: you are accountable for delivering safe, quality care regardless of whether AI was involved.
Does AHPRA's guidance create new AI rules?
No. The most common misunderstanding is that the guidance created a new set of AI rules. It confirmed that existing obligations already cover AI. A note drafted by an AI scribe is your note. A decision informed by an AI tool is your decision. The Code of Conduct does the regulatory work.
That changes how a health group should approach adoption. The question becomes "can we meet our existing professional obligations while using this?" instead of "is this tool allowed?" It is a higher and more useful bar, because it focuses attention on human review, honest disclosure, understanding the tool's limits and keeping patient data safe.
| What practices often assume | What AHPRA's guidance says | Impact | |
|---|---|---|---|
| Legal basis | AI needs a new special rule | Existing Code of Conduct already applies | Clarified |
| Responsibility | The vendor or model is responsible | The practitioner remains accountable | Unchanged |
| AI output | Can be trusted if it looks right | Must have human judgement applied | Verified |
| Patients | Do not need to know | Should be told AI is being used | Transparent |
| The tool | Use it, do not need the detail | Understand its training, use and limits | Informed |
How do AHPRA's obligations apply to AI scribes?
For most practices the first and biggest AI use case is the ambient scribe, which listens to the consultation and produces a draft clinical note or letter. It is also where AHPRA's principles bite hardest, because the output becomes a clinical record.
Under the accountability principle, the draft stays a draft. The practitioner reads what the scribe produced, corrects errors and omissions, and confirms the note reflects what happened in the room before it enters the record. AI scribes can occasionally invent detail that was never said, misattribute symptoms, or miss something significant mentioned in passing. Signing an unread AI note is a professional risk.
- ConsentPatient informed AI is used
- CaptureScribe drafts the note
- ReviewClinician checks and corrects
- SignPractitioner confirms the record
The workflow is simple. The compliance work lies in never letting the middle steps collapse on a busy day. Make the human review a fixed step, however good the tool gets.
What does transparency and informed consent mean for AI in healthcare?
AHPRA is explicit that patients should be informed when AI is used in their care and that their concerns should be addressed. An AI scribe captures a patient's voice and health information, and they are entitled to know.
Transparency does not need a legal lecture at every appointment. It needs a clear, standing way of informing patients, together with a real option to decline:
- a notice in the waiting room and on your website
- a line in your patient information
- a plain verbal mention before recording begins
- a workable fallback to traditional documentation if a patient declines
Consent ties directly to privacy law. Health information is sensitive information under the Privacy Act 1988, and recording a consultation and sending it to an AI tool is a collection and disclosure of that information. Our comparison of the Australian Privacy Act and the GDPR is worth reading before you sign a scribe contract.
How much must practitioners understand about their AI tools?
Understanding is the principle practices most often overlook. AHPRA expects practitioners to understand enough about the AI they use, including how it is trained and tested, its intended use and its limitations. You do not need to be a machine learning engineer. You do need to answer: what was this tool built to do, what is it unreliable at, where does the data go, and how do we know it works for our patient population?
Can you answer yes to all of these?
- You understand its intended use and limitationsNecessary before clinical use
- Patient data stays protected and appropriately locatedConfirm before any deployment
- A human reviews every clinically relevant outputNon-negotiable
- Patients are informed and can declineBuild it into the workflow
That makes vendor selection a clinical governance issue as well as a procurement one. Our AI vendor selection questions set out what to ask. For a clinical tool, answers about training data, testing, error rates and data handling matter more than the demo. If a tool's intended purpose moves into diagnosis or treatment recommendations, TGA medical device rules may also apply, covered in our guide to AI medical software and the TGA.
Where should patient data from AI tools be stored?
Health information is among the most sensitive data any Australian organisation holds. The OAIC's Notifiable Data Breaches reports have repeatedly listed health service providers among the sectors reporting the most breaches. An AI tool that captures consultations adds a new place that data flows and a new party who handles it.
Answer two questions before any clinical AI tool goes live:
- Where do the audio and text go, and are they stored or processed offshore? For patient data, Australian hosting and clear data-handling terms are part of meeting your privacy obligations. Our data sovereignty guide explains the reasoning.
- Is consultation data used to train the vendor's models? If patient information feeds a shared training process, you have a disclosure problem no reception notice was written to cover. The general failure mode is described in how the wrong AI tools leak business data. In a clinical setting the consequences are sharper.
For health groups that want certainty, AI can run on your own infrastructure in Australia (your cloud or your own servers), so recordings and notes never leave your environment. Our private AI infrastructure work covers that option.
How should a health group roll out AI tools?
Groups that adopt AI well pilot first, write down the rules and train the team before scaling across sites. A staged rollout also creates the evidence trail that shows you took your obligations seriously.
- Weeks 1-2AssessVendor due diligence, data handling, privacy review
- Weeks 3-4PilotA few clinicians, close review of every note
- Weeks 5-6DocumentConsent process, review rule, patient notices
- Week 7+ScaleRoll out with training and ongoing audit
- Clinician documentation time returned per consultReal
- Professional standing under AHPRA obligationsProtected
- Patient trust through transparency and consentMaintained
- Privacy exposure from mishandled health dataReduced
- Accountability kept with the practitionerAlways
The values are qualitative on purpose. Time returned depends on patient mix, specialty and prior documentation load. Measure it in your own clinics instead of trusting a vendor's headline figure, and weigh it against the review discipline the tool requires.
How does AHPRA's guidance fit wider practice automation?
The same governance thinking applies whether a tool drafts notes, handles patient messages or automates back-office work. Our guide to AI patient automation for healthcare practices covers the operational side, and pharmacies face parallel questions covered in AI for pharmacies.
Underneath sits one requirement: a governance structure that decides who approves a tool, what data it can touch and who signs off its outputs. Our AI governance framework guide sets out that structure, and the AI ethics framework covers the values questions that matter in a care setting.
AI drafts, listens and lightens the load. It does not carry the duty of care, does not decide, and does not replace the judgement a practitioner's registration represents.
What should health groups do this quarter?
- Inventory your AI. List every tool that touches patient care or patient information, including ones individual clinicians adopted on their own.
- Write the review rule down. No AI-generated clinical note or letter enters the record without a practitioner reading, correcting and confirming it.
- Fix your consent process. Put patient notices in place and give clinicians a simple way to inform patients and offer a fallback.
- Interrogate data handling. Confirm where patient data is processed and stored, whether it trains models, and whether hosting meets your privacy obligations.
- Document your diligence. Keep a short record of each tool, its intended use and limits, and the checks you did. That trail is your evidence.
How Solve8 can help
Our team helps healthcare and allied health groups inventory their AI tools, check data handling and build governed workflows where a clinician approves before anything reaches the record. Where data residency matters, we can run AI on your infrastructure in Australia. See our AI strategy services or talk to us.
Sources: AHPRA and National Boards guidance, "Meeting your professional obligations when using Artificial Intelligence in healthcare"; the relevant National Board Codes of Conduct; Privacy Act 1988 and OAIC guidance on health and sensitive information and the Notifiable Data Breaches scheme. This article is general information, not legal or clinical advice.
