Police check and health exam emails and AI data leakage: what a document checklist gives away in ChatGPT or Claude.

Once a visa application is ready, the agent tells the client which police certificates and health examinations are needed and when. The email is a checklist, but a checklist for a real family names the conviction the Chilean certificate will show, the child whose asthma the medical will record and the appointment times. Staff paste it into ChatGPT or Claude to make it clearer and kinder. This page marks every item in a routine checklist email that the Code of Conduct and the Privacy Act bear on, and shows how the same email is drafted on Ermos inside the practice.

The police check and health examination email

Sent after the application is prepared and before lodgement or in response to a Home Affairs request. Home Affairs may ask for health examinations and gives each person a HAP ID. Section 39 of the Code requires an agent to advise a client in writing when an application is lodged and to keep the client informed of material developments.

To:
From:
Date: 2 November 2026
Subject: What you need before we can lodge

Dear Camila and Sebastián,

Police certificates.

Health examinations. Home Affairs gives each person a HAP ID when it asks for examinations.

Appointments.

Kind regards,

An invented email. The family, the doctor and the HAP IDs are invented or placeholders. Hover or tap any highlighted item for the analysis; the full list is below.

What leaves the practice

  1. 1
    The clients

    Two named adults. Personal information, and the key that ties the rest of the email to a real family. Privacy Act guide.

  2. 2
    A child’s health condition and her doctor

    Asthma, an action plan and a named doctor. Health information is sensitive information where the Privacy Act applies, and it feeds the health requirement for the visa. Patient data guide.

  3. 3
    A conviction and what the certificate will show

    A named person’s drink-driving conviction and the practice’s advice about it. A criminal record is sensitive information under the Privacy Act, and the practice’s approach to declaring it is advice to the client. Privacy Act guide.

  4. 4
    HAP IDs for each family member

    Home Affairs identifiers linked to named people, including two children. Placeholders here, but a real HAP ID is used to book examinations and is tied to a person. Privacy Act guide.

  5. 5
    The rule about police certificates

    Home Affairs asks for certificates from countries where a person has lived for 12 months or more in the past 10 years since turning 16, and the pages differ by visa. A public rule, with no client data in it.

  6. 6
    The visa expiry and the validity of results

    The date her visa ends and the usual 12-month validity of health examination results. The first is the client’s deadline, the second is public information. Privacy Act guide.

  7. 7
    Appointment times and the children’s schooling

    When and where four people will be at the clinic and that the children will be out of school. Personal information, and a timetable for the family. Privacy Act guide.

  8. 8
    The sender

    A named staff member and an agent’s MARN. Ties the email to a practice bound by the Code of Conduct.

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 client file. 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 migration practices.

Common questions

Do ChatGPT and Claude use the information in a police check and health examination email?
Yes, whatever is pasted from the 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. The email names four HAP IDs, a child's asthma and her doctor, a 2017 drink-driving conviction and the clinic appointment times.
Is it safe to paste a police check and health examination email into ChatGPT or Claude if we remove the names?
No, because the details still point to one family. 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. Four HAP IDs, a Chilean certificate that will show a 2017 conviction, a named doctor and clinic times on 17 November fit only one household. 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 a ChatGPT Business or Claude Team plan make it safe for a police check and health examination email?
Not on its own. A business plan changes the default on training, but it does not change who receives the content. On a personal plan the user cannot choose where conversations are stored or processed. Some business plans let an organisation choose to store data at rest in Australia (OpenAI), which still leaves the content with the AI company. The OAIC's guidance also tells organisations to review carefully the terms and settings that will apply to their use of any AI product. This email names a child's diagnosis and an adult's conviction, so both would still reach the AI company in full. Our comparisons with ChatGPT Enterprise and Claude weigh the trade-offs.
How can a migration practice use AI for a police check and health examination email 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 email is drafted from the client file and the practice's own checklist template. The registered migration agent checks each requirement against the visa's own checklist before approving it.
Do police certificate and health examination requirements differ by visa?
Yes. Home Affairs pages describe the requirement in slightly different terms, for example by age and by how time is counted, and health examinations are requested through a HAP ID. The visa's own checklist or the department's request is the final word, so an agent should confirm the current rule for the subclass rather than rely on a general summary.
Why is a health appointment email sensitive if no results are in it?
Because it names who is being examined, when and why, and may refer to a diagnosis. Health information is sensitive information under the Privacy Act where the Act applies to the practice. The Code of Conduct's duty in section 35 applies in any case, and covers personal information relating to a client's affairs.
Staff want AI to make checklist emails clearer. Is there a safer way?
Use a template with placeholders, or draft where the data stays. On an Ermos unit the email is built from the file on hardware in the office. On Edge it stays on the practice's own instance in Australia. The agent still checks the rule against the visa's checklist before the email goes out.

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 migration practices Resources The Privacy Act, the APPs and AI All professions