Repeat script requests and AI patient privacy: what a repeat prescription email gives away in ChatGPT or Claude.

Repeat prescription requests arrive by email, portal message and phone every day. Reception staff answer them from templates, and when a message is long, emotional or awkward they paste it into ChatGPT or Claude to draft a polite reply. The message names medicines that say exactly what the patient is being treated for, and patients often add more than they would say at the desk. This page marks a routine request and shows how the reply is drafted on Ermos inside the practice.

The repeat prescription request

Sent by a patient to the practice inbox or patient portal. Reception triages it, a GP decides on the prescription, and staff reply to the patient.

Email to reception@wombeyanstfp.example.com.au
Monday 2 November 2026, 7:41 am

Subject: repeat scripts please

Hi, could Dr Vasilakis please send through repeats for and ?

Reception draft reply

An invented email. The email domain is a reserved example domain and the mobile number is masked. Hover or tap any highlighted item for the analysis; the full list is below.

What leaves the practice

  1. 1
    The patient’s name, date of birth and mobile

    Identifies him and how to reach him. Personal information collected in providing a health service, so held as health information under section 6FA. Privacy Act guide.

  2. 2
    An opioid

    Oxycodone is a Schedule 8 medicine. Its name against a patient says he is treated for significant pain, and the record of it is health information. Patient data guide.

  3. 3
    An antipsychotic

    A medicine name that points to a mental health condition. Health information is sensitive information under the Privacy Act. Patient data guide.

  4. 4
    Pharmacy and a relative collecting

    A named pharmacy, a named brother and the fact that he is not driving. Privacy Act guide.

  5. 5
    A relapse disclosed in writing

    Drinking again after a separation, written because he could not say it in person. The kind of confidence Good medical practice 4.4 asks doctors and their staff to keep. Patient data guide.

  6. 6
    An instruction about an ex-partner

    A request that someone specific is told nothing. The practice has to be able to honour it, which it cannot do for a copy held elsewhere. Privacy Act guide.

  7. 7
    The sign-off

    A first name with nothing new in it.

  8. 8
    The reception draft

    Standard wording from the practice template. Nothing in it is about his health.

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 EdgeErmos Dominion
Where the document is processedProvider servers, generally offshore on personal plansYour instance in an Australian data centre; answers generated in AustraliaThe unit in your office
Who can read itProvider staff and systems, under policyYour staff, within the boundaries you set. The named Australian provider processes only permitted excerpts, under contract, and keeps nothingYour staff only
Is it keptOn the provider’s termsKept on your instance; the provider is barred from retaining itStays on the unit
Can it train a modelOn consumer tiers, yesNo, by contractOnly your own unit learns from it, on your hardware

Part of AI data leakage 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 repeat prescription request?
Yes, a pasted patient email is sent to the AI company's servers and processed there to produce a reply. As at October 2026, whether a conversation on a personal plan (ChatGPT Free, Plus or Pro; Claude Free, Pro or Max) can be used to train the models depends on a setting each user controls. Business plans (ChatGPT Business and Enterprise, Claude Team and Enterprise) and the developer APIs sit under commercial terms, and OpenAI and Anthropic each say they do not train on that content by default (OpenAI, Anthropic). On either product, the thumbs up and thumbs down buttons can send a whole conversation back for training. This email names an opioid and an antipsychotic, a pharmacy, a brother collecting the script, a relapse into drinking and an instruction about an ex-wife.
Is it safe to paste a repeat prescription request into ChatGPT or Claude if we remove the name?
No, because the medicines and the story still identify him. Removing names rarely makes a document anonymous. Information stays personal information while the person is reasonably identifiable, and the OAIC warns that information can be at risk of re-identification even when it has been de-identified or anonymised. A named pharmacy, a brother called Kerem and a recent separation point to one patient. The Office of the Australian Information Commissioner (OAIC) recommends in its guidance on commercially available AI products that organisations do not enter personal information, and particularly sensitive information, into publicly available generative AI tools, because of the significant and complex privacy risks.
Does deleting the chat remove a repeat prescription request from ChatGPT or Claude?
Not completely. Deleting a chat removes it from the user's history, but OpenAI and Anthropic each say they can keep a deleted chat on their systems for up to 30 days, and longer where safety or legal obligations apply (OpenAI, Anthropic). Both products also have memory features that can carry details from one conversation into later ones, and on a personal account each user decides whether they are switched on (ChatGPT, Claude). A deleted chat that held this email would still have held a patient's relapse and his request that his ex-wife is told nothing, and the practice could not tell him the message had gone.
How can a GP use AI for repeat prescription requests without sending it to ChatGPT or Claude?
Ermos is built so the document stays under your control. On the Dominion AI unit it is drafted inside your office and none of it is sent to an AI provider. On Ermos Edge your documents and index stay on your own instance in an Australian data centre, and only the question and the excerpts the user is permitted to see go to a named Australian provider that is contractually barred from keeping them or training any model on them. Replies are drafted from the practice's own templates, and any question about medicine goes to a GP. Reception checks each reply before it is sent.
Can reception staff paste patient emails into ChatGPT to write replies?
Doing so sends the patient's message, including medicine names and anything else they wrote, to the AI provider. The practice has to be able to justify that disclosure under APP 6, and the OAIC recommends against entering sensitive information into publicly available AI tools.
Is a medicine name health information on its own?
Against a named patient, yes. A medicine name usually reveals the condition being treated, and it was collected in providing a health service.
Who decides whether to issue the repeat?
The GP, in the practice's clinical software. Reception drafts the reply only after that decision is made.

See Dominion answering from your own documents.

A demonstration takes 30 minutes, uses no client data, and comes with no obligation.

Book a demo