NDIS Provider Participant Emergency and Disaster Plans: Use case and AI Confidentiality

Registered NDIS providers of supports such as daily personal activities and community participation have to plan for emergencies and disasters, and the NDIS Commission’s guidance says personal emergency plans should be developed with the participant’s input and communicated to the people who support them. Before summer, team leaders paste these plans into ChatGPT or Claude to update them, make an Easy Read version or check nothing is missing. A personal plan is a complete picture of a vulnerable person: where he lives alone, his diagnosis, when his injection is due, what he looks like when he is unwell, who to call and how to get him out. This page marks every item in one NDIS emergency and disaster plan that the Practice Standards and the Privacy Act bear on, and shows how the same plan is updated on Ermos inside the service.

The participant emergency and disaster plan

Written with the participant by the team leader, shared with his workers and support network, and reviewed before summer and whenever his circumstances change. The Emergency and Disaster Management standard requires planning that considers and mitigates risks to participants in an emergency or disaster and keeps critical supports going. The Commission’s December 2024 provider alert says personal plans should be reviewed regularly and every time a change may affect a person’s ability to evacuate.

Brushbox Community Supports · Personal emergency and disaster plan
Reviewed with Sam at home on Friday 13 November 2026 by Beth Hargreaves, Team Leader · Copies on his fridge and in his grab bag

About Sam
Health and medication
When stress builds
Who to call
Heat and power
Leaving home
If we cannot staff a shift
Sharing
Everyone
Practice and review

An invented plan. The NDIS number and phone numbers are placeholders. Hover or tap any highlighted item for the analysis; the full list is below.

Emergency plan drafter

An Agent+ that drafts a participant's emergency and disaster plan from their support plan, risk assessment and your plan template. It places medication notes, living arrangements and contacts in the right sections, and notes anyone the participant has asked not to be told, for the team leader to confirm.

What it does

  1. Reads the support plan and risk assessment
  2. Drafts each section from your plan template
  3. Flags missing contacts, supports and unplanned risks

What stays with your people. The team leader completes the plan with the participant and reviews it when circumstances change. The agent drafts; it makes no planning decisions.

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 service

  1. 1
    Sam, his home and his week

    A named man who lives alone, his address, his NDIS number, his phone and the mornings a worker is with him. Personal information, and a map of when he is on his own. The Commission’s alert names living alone and having few informal supports among the situations that put people with disability at greater risk in an emergency. Privacy Act guide.

  2. 2
    Diagnosis, injection dates and medication

    A mental illness, the clinic, the date of his next injection, a second condition and where his tablets are kept. Health information, which is sensitive information under the Privacy Act. The Commission’s alert says a personal plan should include important health information and medication. Patient data guide.

  3. 3
    What relapse looks like and what helps

    The signs that he is becoming unwell and how to support him. The alert lists strategies for supporting mental health among what a personal plan should hold. Health information, and the paragraph most likely to be used against him if it surfaced. Patient data guide.

  4. 4
    His contacts

    A cousin’s name and number, his GP and his clinic. Third parties’ personal information, and a list of who would notice if something happened to him. Privacy Act guide.

  5. 5
    Heat, power and fire risk

    No air conditioning, where the torch is, and that he smokes and once burnt a cushion. Details about his home that bear on his physical safety. The Emergency and Disaster Management standard requires planning that considers and mitigates risks to participants’ health, safety and wellbeing in an emergency or disaster.

  6. 6
    How he leaves home

    Where he goes, why crowds are hard for him and what is in his grab bag. The alert says a personal plan should cover different scenarios, such as when to seek shelter, going to an evacuation point and losing power. Health information sits inside the logistics. Patient data guide.

  7. 7
    Continuity of critical supports

    Which supports cannot lapse and what the service does if it has no worker, with the names of the three people who know him. The quality indicators expect measures that keep supports critical to each participant’s safety, health and wellbeing going before, during and after an emergency or disaster. Operational detail about a small service as well as about him.

  8. 8
    Who may and may not be told

    His agreement to share the plan with named people, and a request to keep his diagnosis from his agent and neighbours. The alert says plans need to be communicated to a person’s support network and workers, and the Information Management indicators expect consent to cover the purpose of a disclosure. An AI company is on neither list. Privacy Act guide.

  9. 9
    The general emergency steps

    Call 000, follow emergency services, then the on-call team leader. Wording the service puts in every plan, naming no one.

  10. 10
    Rehearsal and review

    The date he practised leaving, how long it took and when the plan is next reviewed. The alert says plans should be reviewed regularly and whenever a change may affect a person’s ability to evacuate, and drills practised routinely. Low risk alone; the review date is a commitment the service has to keep.

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 participant 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 Use cases in NDIS services.

Common questions

Do ChatGPT and Claude use the information in a personal emergency and disaster plan?
Everything pasted or uploaded goes to the AI company's servers and is 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. A plan pasted in for an Easy Read version carries Sam Ibrahim's address and the mornings a worker is with him. It also gives the date of his next depot injection and the signs that he is becoming unwell.
Can we use ChatGPT or Claude for an emergency plan if we remove the names?
No, because the plan is written to pick out one man in one home. 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 46-year-old living alone in a ground-floor unit in Gateshead, with no car, is easy for his neighbours to place. A depot injection every second Thursday at the community mental health clinic points to him as well. His schizophrenia and diabetes are health information, which is sensitive information under the Privacy Act. Sam has also asked that his neighbours and his real estate agent are not told his diagnosis. 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 an emergency and disaster plan from ChatGPT or Claude?
Not in a way the service could show Sam. 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). The service can name every copy of his plan, on his fridge, in his grab bag and with the workers, Nadia and Dr Sefton he agreed to. Pasting it adds a recipient Sam never named. After the chat is deleted, the service still has no way to tell him that his injection dates and relapse signs have gone.
How can an NDIS provider use AI for a personal emergency and disaster plan 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 updated plan or Easy Read copy is drafted from Sam's file and the service's own plan template. Beth reviews it with Sam, checks the contacts, medication and evacuation steps, and signs it before new copies go on his fridge and in his grab bag.
Do registered NDIS providers need emergency and disaster plans?
Registered providers assessed against the core module of the Practice Standards do. The core module covers supports such as assistance with daily personal activities and community participation; the shorter verification module used for some other supports does not include this standard. The Emergency and Disaster Management standard requires planning that considers and mitigates risks to participants' health, safety and wellbeing in an emergency or disaster, and keeps critical supports going. Its quality indicators expect the governing body to develop the plans, consult participants and their support networks, test and review them, communicate them to workers and participants, and have each worker trained to put them in place.
What should a participant's personal emergency plan include?
The NDIS Commission's December 2024 provider alert lists emergency contacts and the support network, important health information including medication and the health professionals involved, preferred communication, strategies for mental health or behaviour support, mobility or communication aids and medical devices, and, with the participant's permission, other providers. It also says plans should cover scenarios such as sheltering, going to an evacuation point and losing power.
How often should a personal emergency plan be reviewed?
The Commission's alert says personal emergency plans should be reviewed regularly and every time there is a change in circumstances that may affect a person's ability to evacuate, and that drills should be practised routinely. The Support Planning quality indicators also expect each participant's support plan to anticipate emergencies and disasters and to be understood by each worker who supports them.

See Dominion answering from your own documents.

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

Book a demo