Transfer letters and AI security: what a hospital handover gives away in ChatGPT or Claude.

The transfer letter travels with the resident in the ambulance. It has to tell an emergency department everything that matters in one page, which makes it the densest clinical document a provider writes and the one most tempting to have ChatGPT or Claude compress. Liaising with a hospital is an authorised disclosure; drafting the letter in a frontier model is a separate one. This page marks every item in a transfer letter that the Aged Care Act 2024 and the Privacy Act protect, and shows how the same letter is generated on Ermos inside the service.

The hospital transfer letter

Prepared by the registered nurse when a resident is transferred to hospital, and handed to the paramedics.

Transfer to St Hilda’s Private Hospital Emergency Department
From Kurrajong Gardens Aged Care · 24 October 2026, 5.55am · RN

Resident.

Reason for transfer.

Background.

Medications.

Observations at 5.50am.

Advance care directive.

Next of kin.

An invented letter. The Medicare number is a published test value. Hover or tap any highlighted item for the analysis; the full list is below.

What leaves the service

  1. 1
    Identity, Medicare and health fund

    Name, date of birth, Medicare number and private health membership. The identifiers used for health system fraud, about a woman with dementia. Privacy Act guide.

  2. 2
    The reason for transfer

    The injury and the anticoagulant concern. Health information, and the clinical core of the letter. Patient data guide.

  3. 3
    Diagnoses and how to approach her

    Her conditions and the behavioural notes that keep her and staff safe. Sensitive information, written for the hospital’s benefit. Patient data guide.

  4. 4
    The medication list

    Including a psychotropic used for behaviour, which is a restrictive practice under the Act. Patient data guide.

  5. 5
    Observations

    Clinical measurements. Health information. Patient data guide.

  6. 6
    The advance care directive

    Her decisions about resuscitation, intensive care and surgery. The most consequential and personal information in the file. Patient data guide.

  7. 7
    Next of kin and a family restriction

    Two named relatives, a phone number, and an instruction about who may be told. Their personal information and a confidence. Privacy Act guide.

  8. 8
    The nurse

    Ties the letter to a named worker and a direct phone line.

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 service 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 resident’s 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 aged care providers.

Common questions

Do ChatGPT and Claude learn from a hospital transfer letter?
They can. A pasted letter is sent to the AI company and processed there to produce the 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. A transfer letter puts the identifiers, the medication list and the advance care directive on one page.
Is it safe to paste a hospital transfer letter into ChatGPT or Claude?
No, because the letter ties the resident's identity to her diagnoses and medications. 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. The name, date of birth, Medicare number and health fund membership identify the resident. Without them, the diagnoses, the apixaban dose and the end-of-life instructions would still describe one person.
Can ChatGPT or Claude be trusted to draft a hospital transfer letter?
Not without a registered nurse checking every line against the chart before the ambulance leaves. Language models can write fluent text that adds details missing from the source, changes a figure or a date, or cites a result that looks real and was never recorded. The most harmful errors would sit in the medication list, the time of the last apixaban dose, the allergy line, the observations and the advance care directive. An emergency department would read each one as fact.
How can an aged care service use AI for a hospital transfer letter 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. The letter is drafted from the care plan, the medication chart and the morning's observations in the service's own transfer template. The registered nurse checks it, signs it and prints it for the ambulance.
The hospital needs all of this. Isn't the disclosure authorised?
Disclosure to the hospital is for a purpose connected with the resident's care, which section 168 of the Act allows, and the hospital is bound by its own health records obligations. Drafting the letter in a frontier model sends the same information to a company with no role in her care, retained on its terms, which is a separate disclosure nobody authorised.
Transfer letters are urgent. Is there time to do this inside the service?
That is the point of generating it inside the service. On an Ermos unit the letter is produced from the care plan, the medication chart and the morning's observations in the time it takes to ask, and the nurse checks and prints it. Speed is the reason staff reach for a chatbot; a unit in the building is faster.
What about the advance care directive?
It records what the resident decided about her own death while she could decide. It is the paragraph the hospital most needs and the one that should never exist anywhere a stranger could read it.

See Dominion answering from your own documents.

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

Book a demo
Keep reading
AI data leakage in aged care providers AI for healthcare AI and patient data All professions