Two years ago, an AI tool that listened to your consult and wrote the note for you sounded like a pitch deck. Now it is close to standard kit in Australian general practice. The shift was fast and it was real.

The RACGP's newsGP poll put AI scribe use at 22 percent of GPs in August 2024. By November 2025 that figure was 40 percent. Read it carefully, because the question captured any use across a practice. Survey work by Healthed suggests routine personal use sits closer to 20 percent. Either way, the direction is not in doubt.

The money tells the same story. Melbourne-founded Heidi raised a USD $65 million Series B in October 2025, led by Steve Cohen's Point72, taking total funding near USD $96 million. Heidi now says it supports more than two million consultations a week. Its cross-town rival Lyrebird raised USD $12 million in June 2025 and processes around 30,000 consultations a day in Australia alone. PatientNotes and others fill out a crowded field. This is a genuine Australian export story.

What these tools actually do in the room

An ambient scribe records the conversation, then drafts the clinical note from it. The clinician reviews, edits and signs. Lyrebird claims six to eight minutes saved per consult. The doctor still owns every word.

Pricing is modest. PatientNotes sits around AUD $80 to $110 per user each month. Lyrebird is free for Best Practice users, which is a large slice of Australian general practice. At that price, adoption is a clinician's decision long before it reaches an executive's desk. That is the part health leaders keep underestimating.

The Western Australia gap

Here is the uncomfortable bit. The marquee Australian deployments are happening elsewhere. Gold Coast Health validated Lyrebird across its outpatient clinics and reported a 22 percent lift in patient throughput. Heidi runs at scale inside Monash Health and the Queensland Children's Hospital. Ramsay is piloting its own scribe in an Ipswich emergency department.

Scan for a Western Australian flagship and the page is mostly blank. WA Primary Health Alliance operates all three of the state's Primary Health Networks, which is a rare and powerful position. We could not find a public WA scribe program with named results to match Queensland or Victoria. The clinicians are almost certainly already using these tools. The system-level posture has not caught up.

A thin public record is itself a finding. It usually means adoption is happening quietly, one doctor and one credit card at a time, with no shared view of consent or data residency. That is the worst of both worlds. You get the medico-legal exposure without the governance or the bargaining power.

Because the exposure is real. The RACGP's August 2024 guidance is blunt. The GP is liable for errors in the record even when an AI generated them. Consent should be sought for each consult and patients must be able to opt out. The college discourages first-year registrars leaning on scribes before their own clinical reasoning is formed. The Australian Commission on Safety and Quality in Health Care went further in August 2025 with a dedicated safety scenario for ambient scribes. None of this stops adoption. It just means someone has to own it.

Our view, from building clinical AI and advising the people who deploy it, is simple. The technology is ready enough to use and risky enough to govern. A WA health service that waits for a national mandate will inherit whatever shadow setup its staff have already cobbled together. A service that moves now gets to choose the vendor and set its own consent and data rules. One of those is a strategy.

So which WA health service publishes the first scribe rollout with real numbers attached? If you are weighing that decision, or trying to get a grip on what your clinicians are already running, talk to Coterie Health. We will help you turn quiet adoption into a deliberate program.