GP Reception Phone Call Transcripts: Use case and AI Confidentiality
Call recording and AI receptionists are marketed hard to medical practices, and reception staff already paste call transcripts into ChatGPT or Claude to summarise a message, pick the right appointment or write the call note. A front desk call looks routine. In four minutes the caller gives her name, date of birth, Medicare number, a symptom, a medicine, her husband’s test and a money problem. This page marks every item in an invented reception phone call transcript that the Privacy Act, the NSW health records law and the NSW surveillance devices law bear on, and shows how the same call is handled on Ermos inside the practice.
The reception phone call transcript
Recorded and transcribed when a patient rings the front desk to book, ask about a result, collect a script or query an account. Reception reads the transcript to write the booking note and the message for the doctor.
Wombeyan Street Family Practice
8 Wombeyan Street, Eastwood NSW 2122 · (02) 5550 0161
Call transcript (automatic) · Wednesday 18 November 2026, 9:12 am · inbound · 3 min 36 s
00:02 Recorded message: Thank you for calling Wombeyan Street Family Practice. This call is recorded and transcribed so that we can take accurate messages. If you would rather not be recorded, say so now and we will take your message without a recording.1
00:19 Tahnee: Wombeyan Street Family Practice, Tahnee speaking.
00:22 Caller: Hi Tahnee, it’s Sylvia Wetherby. I need to see Dr Vasilakis as soon as I can.
00:29 Tahnee: Of course, Sylvia. Can I confirm your date of birth and address?
00:34 Caller: Eighth of March 1958, 5 Tilbrook Place, Eastwood.2
00:41 Tahnee: And your Medicare number for the booking?
00:45 Caller: 2999 99999 9, and I am number one on the card.3
01:02 Tahnee: Thank you. What is it about, so I can book the right length of appointment?
01:07 Caller: I found a lump in my left breast about a fortnight ago. I keep touching it and I cannot sleep.4
01:24 Tahnee: I am sorry to hear that. I will make it a long appointment. Dr Vasilakis has Friday 20 November at 10:40.5
01:40 Caller: Thank you. And I have run out of my citalopram, 20 milligrams. I have been on it since my anxiety started after Mum died.6
01:58 Tahnee: The doctor will need to see you for that script, so I will put a note on the booking.
02:06 Caller: Can you tell me if Gordon’s scan has come back? He had it on 9 November for his kidneys and Dr Szabo said they would ring, but Gordon never rings anyone back.7
02:25 Tahnee: I am sorry, I cannot talk about another patient’s results unless they have given us permission. I will leave a note for the doctor’s team to ring Gordon himself.8
02:41 Caller: Fair enough. One more thing. I got an account for my September long consult and I cannot pay it all at once. I was made redundant from Tavener Printing in October.9
03:02 Tahnee: I am sorry to hear that. Odette, our practice manager, can set up a payment plan. Can she ring you?
03:10 Caller: Yes, on 04xx xxx 566, after two o’clock please. Do not leave any details on voicemail.10
03:28 Tahnee: Will do. See you Friday at 10:40, Sylvia. Goodbye.
An invented call transcript. The practice, staff and callers are fictional; the Medicare number is a placeholder that fails its check digit and the mobile number is masked. Hover or tap any highlighted item for the analysis; the full list is below.
Call transcript note drafter
An Agent+ that works from the transcript of a call to the front desk. It follows your reception note format, separates the booking, the script query, the result enquiry and any account question, and drafts a message for the doctor that omits unrelated personal detail for reception to confirm.
What it does
- Reads the transcript from start to finish
- Separates each request in your reception note format
- Drafts the doctor's message and flags anything clinical
What stays with your people. Reception checks the note against the call before it is saved, and the GP reads and decides on every clinical message.
An Agent+ is built once and anyone in the practice can run it. It works from your own documents and runs on Ermos Dominion, the AI unit in your office, or on Ermos Edge, your own instance in an Australian data centre.
What leaves the practice
- 1The recording notice
Section 7(1)(b) of the Surveillance Devices Act 2007 (NSW) says a person must not knowingly use a listening device to record a private conversation to which the person is a party, and s 7(3)(a) says that does not apply where all the principal parties consent, expressly or impliedly. Whether a front desk call is a private conversation, and whether continuing after a notice is consent, is a question for advice. APP 5.1 asks for reasonable steps to make an individual aware of the matters in APP 5.2, which include the purposes of collection, who the information is usually disclosed to and whether it is likely to go to overseas recipients. The wording is the practice’s own, with nothing personal in it.
Frontier Models (ChatGPT, Claude) The practice’s own wording, nothing personal Ermos Edge Not about where data goes Ermos Dominion Not about where data goes Use case Call transcript note drafter. Reads the call transcript and drafts the booking note and the message for the doctor, keeping each request separate and marking details that need a clinician. See the use case - 2Name, date of birth and address
The identity check. Other personal information collected in providing a health service is health information (Privacy Act s 6FA(b)), and APP 11.1 asks an organisation holding personal information to take reasonable steps to protect it from misuse, interference and loss and from unauthorised access, modification or disclosure. Privacy Act guide.
Frontier Models (ChatGPT, Claude) Transmitted to the provider, generally offshore on personal plans Ermos Edge Stays on your instance in Australia; leaves only inside a permitted excerpt, per question, to the named provider Ermos Dominion Stays on the unit in your office Use case Call transcript note drafter. Reads the call transcript and drafts the booking note and the message for the doctor, keeping each request separate and marking details that need a clinician. See the use case - 3A Medicare number
An identifier assigned by a government agency. APP 9.2 limits when an organisation may use or disclose a government related identifier. Read aloud on a call, it becomes text in a transcript that anyone with access to the tool can search. Privacy Act guide.
Frontier Models (ChatGPT, Claude) A government related identifier transmitted Ermos Edge Stays on your instance in Australia; leaves only inside a permitted excerpt, per question, to the named provider Ermos Dominion Stays on the unit in your office Use case Call transcript note drafter. Reads the call transcript and drafts the booking note and the message for the doctor, keeping each request separate and marking details that need a clinician. See the use case - 4A symptom and what the caller fears
A breast lump and its effect on her sleep. Health information is sensitive information under s 6(1) of the Privacy Act. APP 3.3 allows an organisation to collect sensitive information with the individual’s consent where it is reasonably necessary for its functions or activities, so the detail asked for should be what booking needs. Patient data guide.
Frontier Models (ChatGPT, Claude) Sensitive health information transmitted; on consumer tiers it can train future models Ermos Edge Stays on your instance; an excerpt holding it could reach the named Australian provider, which keeps nothing; a summary can still swap left and right, so staff check it against the call Ermos Dominion Stays on the unit; a summary can still swap left and right, so staff check it against the call Use case Call transcript note drafter. Reads the call transcript and drafts the booking note and the message for the doctor, keeping each request separate and marking details that need a clinician. See the use case - 5The appointment
A doctor, a date and a time. It links her to one clinician and to a long consultation, which is a hint about the reason.
Frontier Models (ChatGPT, Claude) A doctor, a date and a time; low risk alone, but it links her to the symptom Ermos Edge Not about where data goes Ermos Dominion Not about where data goes Use case Call transcript note drafter. Reads the call transcript and drafts the booking note and the message for the doctor, keeping each request separate and marking details that need a clinician. See the use case - 6A medicine and the history behind it
A named antidepressant and dose, and a bereavement as the reason. Health information, with a family detail attached. Patient data guide.
Frontier Models (ChatGPT, Claude) A medicine, a dose and a bereavement transmitted Ermos Edge Stays on your instance in Australia; leaves only inside a permitted excerpt, per question, to the named provider Ermos Dominion Stays on the unit in your office Use case Call transcript note drafter. Reads the call transcript and drafts the booking note and the message for the doctor, keeping each request separate and marking details that need a clinician. See the use case - 7A husband’s scan
Another patient’s test, its date, its reason and his doctor, volunteered by his wife. He is not on the call and has not agreed to its being recorded or passed on. Good medical practice 4.4.1 asks doctors to treat information about patients as confidential, and APP 6.1 limits using or disclosing it for another purpose. Patient data guide.
Frontier Models (ChatGPT, Claude) Another patient’s test and his doctor transmitted, though he is not on the call Ermos Edge Stays on your instance in Australia; leaves only inside a permitted excerpt, per question, to the named provider Ermos Dominion Stays on the unit in your office Use case Call transcript note drafter. Reads the call transcript and drafts the booking note and the message for the doctor, keeping each request separate and marking details that need a clinician. See the use case - 8The reception reply
Standard wording about not discussing another patient’s results without permission. It contains nothing about any patient and is a practice’s own script.
Frontier Models (ChatGPT, Claude) Standard wording with nothing about any patient Ermos Edge Not about where data goes Ermos Dominion Not about where data goes Use case Call transcript note drafter. Reads the call transcript and drafts the booking note and the message for the doctor, keeping each request separate and marking details that need a clinician. See the use case - 9An account and a redundancy
A billing query, a money problem and a named former employer. This is personal information, held with the patient’s other records, but financial detail is not on the closed list of sensitive information in s 6(1). It still shows why she is being treated differently from other callers.
Frontier Models (ChatGPT, Claude) A billing query, a redundancy and a named employer transmitted Ermos Edge Stays on your instance in Australia; leaves only inside a permitted excerpt, per question, to the named provider Ermos Dominion Stays on the unit in your office Use case Call transcript note drafter. Reads the call transcript and drafts the booking note and the message for the doctor, keeping each request separate and marking details that need a clinician. See the use case - 10A mobile number and a request about voicemail
A masked mobile, a preferred time and an instruction not to leave details. The practice has to be able to follow it, and a transcript that carries it into an outside tool puts the promise outside the practice’s control. Privacy Act guide.
Frontier Models (ChatGPT, Claude) A mobile and a privacy request transmitted Ermos Edge Stays on your instance in Australia; leaves only inside a permitted excerpt, per question, to the named provider Ermos Dominion Stays on the unit in your office Use case Call transcript note drafter. Reads the call transcript and drafts the booking note and the message for the doctor, keeping each request separate and marking details that need a clinician. See the use case
What happens in Frontier Models (ChatGPT, Claude)
- The whole document is transmitted to the provider’s servers. On personal plans the user cannot choose where it is stored or processed, and for the major providers that has generally meant outside Australia, under that country’s legal process. Some business plans offer storage in Australia, but the provider still receives the content.
- On consumer tiers, prompts can be retained and used to train future models.
- The provider’s staff and systems can reach the content under the provider’s own policies.
- Once sent, the practice no longer controls it. Retention and deletion happen on the provider’s terms.
The same document on Ermos
Ermos Dominion
The document is drafted on a unit in your office, from your own templates and the patient record. Every marked item stays on the unit, which keeps working with the internet switched off. Nothing in it is transmitted to an AI provider, so there is no disclosure to analyse: the confidentiality work is the same as for your file server.
Ermos Edge
Your documents, the index and your access boundaries stay on your own instance in an Australian data centre. When a staff member asks for this draft, only their question and the excerpts they are permitted to see go to a named Australian provider, who is contractually barred from keeping the content or training any model on it. Nothing is processed outside Australia.
| Frontier Models (ChatGPT, Claude) | Ermos Edge | Ermos Dominion | |
|---|---|---|---|
| Where the document is processed | Provider servers, generally offshore on personal plans | Your instance in an Australian data centre; answers generated in Australia | The unit in your office |
| Who can read it | Provider staff and systems, under policy | Your staff, within the boundaries you set. The named Australian provider processes only permitted excerpts, under contract, and keeps nothing | Your staff only |
| Is it kept | On the provider’s terms | Kept on your instance; the provider is barred from retaining it | Stays on the unit |
| Can it train a model | On consumer tiers, yes | No, by contract | Only your own unit learns from it, on your hardware |
Part of Use cases in general practitioner (GP) practices. See how this works in practice: Patient messages on Ermos.
Common questions
Do ChatGPT and Claude use the information in a reception call transcript?
Can we use ChatGPT or Claude for a reception call transcript if we remove the names?
What happens to a reception call when an AI transcription tool records it?
How can a GP practice use AI for reception call notes without sending the call to ChatGPT or Claude?
Can a practice record reception calls in NSW?
What must a practice tell callers when it records or transcribes calls?
Can staff paste a call transcript into ChatGPT or Claude to write the message?
See Dominion answering from your own documents.
A demonstration takes 30 minutes, uses no client data, and comes with no obligation.
Book a demo