Falls risk assessments and AI leakage: what a post-fall review gives away in ChatGPT or Claude.

After a fall, a nurse completes a falls risk assessment and reviews what happened. It is a form, so it looks easy to hand to ChatGPT or Claude to complete from the notes or to turn into a summary for the doctor and the family. It holds the fall and the injury, the medicines that made it more likely, how the resident thinks and moves, what the service has put in place and a risk score. Some of it feeds the service’s reporting. This page marks every item in a realistic post-fall assessment and shows what the Aged Care Act 2024 and the Privacy Act bear on, and how the same form is completed on Ermos.

The falls risk assessment

Completed by the registered nurse after a fall, at admission and on any change in condition, and reviewed with the doctor and the physiotherapist.

Kurrajong Gardens Aged Care
Falls risk assessment and post-fall review · 25 October 2026

Resident
Fall being reviewed
Medicines that raise risk
Cognition and behaviour
Physical assessment
What is in place
Risk rating and record
Family and decision-maker
Assessed by
Domains assessed

An invented form and score. Hover or tap any highlighted item for the analysis; the full list is below.

What leaves the service

  1. 1
    The resident

    A named woman, her room and her date of birth. Personal information that sections 23(5) and 168 of the Aged Care Act 2024 require the provider to protect. Privacy Act guide.

  2. 2
    The fall, the injury and the earlier falls

    Three falls with dates, places and injuries, and a hospital attendance. Health information. A fall is not on its own one of the reportable incident types in section 16 of the Act, but it can become one if it involves a listed type such as neglect, so the record needs to be accurate. Patient data guide.

  3. 3
    Medicines that raise risk

    An anticoagulant and a psychotropic. A medicine given to influence behaviour can be chemical restraint, one of the restrictive practice types in rule 17-5 of the Aged Care Rules 2025, with its own consent and review requirements. Health information. Patient data guide.

  4. 4
    Cognition and behaviour

    How she thinks, what she does at night and who she is looking for. Health information, and a husband’s name that says more about her than a diagnosis. Patient data guide.

  5. 5
    Physical findings

    Blood pressure lying and standing, a fracture history, an eye condition and her footwear. Clinical measurements are health information, and a standing drop in blood pressure is the kind of finding a clinician has to check and a draft should not alter. Patient data guide.

  6. 6
    What the service has put in place

    The equipment and routines around her, and a statement that bed rails are not used. Equipment that limits a person’s movement can be a restrictive practice, so the record of what is in place and why matters to the service’s own compliance. Patient data guide.

  7. 7
    The risk score and the reporting trail

    A rating and a note that the fall counts in the quarter’s figures. Residential providers report to the National Aged Care Quality Indicator Program each quarter, including falls and major injury, which the Department of Health guides say is compiled from the care records. A score is a clinician’s judgement and the count is the service’s own data.

  8. 8
    The family decision-maker and a restriction

    A guardian’s name, the time she was called and an instruction that a brother is not to be told. Two people’s personal information and a family confidence. Privacy Act guide.

  9. 9
    The assessor and the doctor

    Names and dates for a nurse, a doctor and a physiotherapy review. Each has professional obligations, and the leak would be traceable to the nurse. Privacy Act guide.

  10. 10
    The assessment headings

    The standard domains on a falls assessment form. Generic, the same on every resident’s form, and safe to share.

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 use the information in a falls risk assessment?
Everything a nurse pastes or uploads 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 assessment holds three falls with dates, Agnes Thornbury's apixaban and risperidone, a standing blood pressure of 96/60 and a risk score of 19 of 25.
Is it safe to paste a falls risk assessment into ChatGPT or Claude if we remove the resident's name?
No, because the assessment still describes her. 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. Three dated falls, a hip fracture repaired in 2024, a husband named Ron and a daughter who is her enduring guardian point to one woman. 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 falls risk assessment?
No. 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. The service must still protect the resident's personal information under section 168 of the Aged Care Act 2024, and a post-fall assessment sends the AI company the fall, the medicines and the risk score. Our comparisons with ChatGPT Enterprise and Claude weigh the trade-offs.
How can an aged care service use AI for a falls risk assessment 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. Ermos prepares the post-fall review from the resident's file and the service's own falls assessment form. The registered nurse checks every reading, medicine and score and signs it.
Is a fall a reportable incident under the Aged Care Act 2024?
Section 16 lists the reportable incident types, and a fall is not one of them on its own. A fall can become reportable if it involves a listed type, such as neglect or unreasonable use of force. The Commission's guidance requires Priority 1 incidents to be notified within 24 hours and Priority 2 within 30 days. Every fall still belongs in the service's own incident system.
Do falls feed into anything the service has to report?
Yes. Residential providers report quarterly to the National Aged Care Quality Indicator Program, which includes a falls and major injury indicator. The Department of Health's guides say the data comes from reviewing each resident's care records for the quarter, so the quality of the falls assessment and the post-fall record affects what the service reports.
Does the small business exemption apply to a small aged care home?
Not in a way that helps here. The Privacy Act treats an activity in the course of providing aged care as a health service, and the OAIC says an organisation providing a health service and holding health information is covered even if it is a small business. A falls assessment is health information, so the Act applies whatever the home's turnover.

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