Behaviour support plans and AI data security: what an NDIS positive behaviour support plan gives away in ChatGPT or Claude.

A behaviour support plan with a restrictive practice is where NDIS law reaches furthest into a participant’s life. In February 2026 the NDIS Commission published a position statement on AI in developing these plans. It does not endorse AI tools, expects participant information to be de-identified if an NDIS provider uses them, and says disclosing sensitive information to a third-party platform such as ChatGPT without safeguards may breach the Code of Conduct. This page marks every item in a plan that includes an environmental restraint and a PRN medication, and shows how the same plan is maintained on Ermos inside the service.

The positive behaviour support plan

Developed by an NDIS behaviour support practitioner with the participant, family and workers, and lodged with the Commissioner when it contains a regulated restrictive practice. The service that implements it reports to the Commissioner monthly and keeps written records of each use. A comprehensive plan is reviewed at least every 12 months.

Comprehensive behaviour support plan
· Version 1, 5 October 2026 · Practitioner · Implemented by Brushbox Community Supports

About Graham.

Behaviours of concern.

What the behaviour does for him.

Proactive strategies.

Restrictive practices.

Authorisation and consent.

How this plan was developed.

Monitoring.

An invented plan. The authorisation reference is a placeholder. Hover or tap any highlighted item for the analysis; the full list is below.

What leaves the service

  1. 1
    The participant

    A named man and where he lives alone. Personal information the Information Management standard expects to be kept confidential. Privacy Act guide.

  2. 2
    Diagnosis and history

    A brain injury, its effects and a history of alcohol-related aggression. Health information, which is sensitive information under the Privacy Act. Patient data guide.

  3. 3
    Behaviours of concern

    Threats, thrown objects, an injury to a worker and leaving the flat after dark. Information that could characterise him unfairly if it surfaced, and information about a worker who was hurt. Patient data guide.

  4. 4
    The function of the behaviour

    What drives his anger, including contact with his ex-wife, named. A third party’s personal information in a clinical document, and the practitioner’s analysis of the man. Privacy Act guide.

  5. 5
    Proactive strategies

    The positive supports the practitioner built with him. The Commission says plans must be person-centred, evidence-informed, uphold the person’s dignity and right to privacy, and be developed in consultation with the participant, family and workers.

  6. 6
    The restrictive practices

    An environmental restraint and a PRN medication. A regulated restrictive practice includes environmental restraint, which restricts free access to items or activities, and chemical restraint, which is medication used for the primary purpose of influencing behaviour. The most regulated decision in the file. Patient data guide.

  7. 7
    Authorisation and consent

    The authorisation reference, the dates and who was consulted. A restrictive practice used other than in accordance with a State or Territory authorisation is a reportable incident. Privacy Act guide.

  8. 8
    The practitioner and how the plan was developed

    A named practitioner, the assessment and the consultation dates. The Commission does not endorse AI tools for developing these plans. An NDIS behaviour support practitioner develops a plan that contains a regulated restrictive practice, and the plan is lodged with the Commissioner.

  9. 9
    Monitoring and review rules

    Each use is recorded, reported to the Commission monthly and the plan reviewed at least every 12 months. The Rules say so for registered providers that use a regulated restrictive practice and for specialist behaviour support providers.

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 AI data leakage in NDIS services.

Common questions

Do ChatGPT and Claude use the information in a positive behaviour support plan?
Yes. A plan pasted into either tool 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 plan carries Graham's brain injury and history of alcohol-related aggression, an injury to a worker, his ex-wife's name and a locked alcohol cupboard alongside a PRN medication.
Can we use ChatGPT or Claude for a behaviour support plan if we remove the names?
Not safely. The Commission's February 2026 position statement endorses or approves no AI tool and expects an NDIS provider using AI to de-identify all information so that no participant's personal information is disclosed to AI systems. 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 brain injury from a 2014 road crash, a locked alcohol cupboard and a named practitioner would still point to Graham. 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.
Can ChatGPT or Claude be trusted to draft a behaviour support plan?
No. Language models can produce fluent text with details that are not in the source, wrong figures, wrong dates or references that look real but are not. An NDIS behaviour support practitioner develops a plan that contains a regulated restrictive practice, and a qualified person should check every item against the file. A wrong authorisation reference or date, a misstated PRN condition or an invented behaviour would do the most harm.
How can an NDIS provider use AI for a behaviour support 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. Any summary or update is produced from the plan on the participant file and the service's own template. The NDIS behaviour support practitioner checks and signs it. The position statement endorses no tool, so the service decides with its own advice whether an arrangement meets its expectation.
What counts as a regulated restrictive practice?
Under the Restrictive Practices and Behaviour Support Rules, five practices are regulated: seclusion; chemical restraint, which is medication used for the primary purpose of influencing behaviour; mechanical restraint; physical restraint; and environmental restraint, which restricts a person's free access to all parts of their environment, including items or activities. Medication prescribed to treat a diagnosed mental disorder or physical condition is excluded from chemical restraint.
Do behaviour support plans have to be lodged, and who reports use?
A plan that contains a regulated restrictive practice, including a reviewed plan, must be lodged with the Commissioner by the specialist behaviour support provider that developed it, as soon as practicable after it is developed, whether or not State or Territory authorisation is required. A registered provider that uses a regulated restrictive practice must give monthly reports to the Commissioner and keep written records of each use.
What did the NDIS Commission say about AI and behaviour support plans?
Its February 2026 position statement says the framework does not explicitly prohibit AI, the Commission does not endorse AI tools, and NDIS providers that use them are expected to de-identify all information so no participant's personal information is disclosed to AI systems. It adds that sending sensitive personal information to a third-party platform such as ChatGPT without safeguards may breach section 6(b) of the Code of Conduct.

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 NDIS services Resources The Privacy Act, the APPs and AI All professions