How we use AI
Artificial intelligence can help with some of the work behind your support. It does not make decisions about you, and your information does not go into it. This page explains where the line sits, and who to tell if you think it has been crossed.
Download the full policyOur rules
From MMT Care's Artificial Intelligence (AI) use policy, version 2.
Four rules sit above everything else. They apply to every worker at every level of the organisation, and to every AI tool — including ones that have not been invented yet.
A person always makes the decision
Human decision-making stays at the core of all supports, services and communications. AI complements our team — it never replaces human experience, expertise and oversight.
Your information stays out of it
Information that is not freely and publicly available is never put into an AI tool without approval — and that includes any participant information and any worker information.
A person checks the output
Anything an AI tool produces is reviewed, quality checked and corrected by a person before it is used. Nothing goes out unread.
We tell you when we use it
We communicate clearly with participants and stakeholders about when and how AI tools are used in supports and interactions. If AI would be used with your information, we ask you first.
What never goes into an AI tool
Information that is not freely and publicly available must never be inputted into any AI engines without prior authorisation from key management personnel.
That covers:
- Any participant information.
- Any worker information.
- Organisational information that is not public — finances, internal correspondence.
Putting something into an AI tool that should not have gone there is treated as a data breach. It is reported, contained and investigated through our incident management process, and where the law requires it, the Office of the Australian Information Commissioner and the people affected are told.
Where we may use it
These uses are approved, as long as the four rules above are followed in full.
- Improving accessibility, in line with the NDIS framework for AI-enabled assistive technologies.
- Worker training, for example AI-powered training simulations.
- Looking at anonymised data to spot broad trends, so we can improve the quality of our services.
- General brainstorming — for example, a list of ideas for continuous improvement.
- Drafting general, non-sensitive documents such as report templates, which a person then reviews and edits.
What we never use it for
These are prohibited outright.
- Diagnosing health conditions.
- Making decisions or creating output — a care plan, a policy — without human review.
- Predicting a participant’s behaviour, progress or outcomes.
- Putting participant or organisational data into any AI tool without approval from management and the participant’s consent.
Your choices
If we ever intend to use AI together with information about you, we have to say so and ask your permission first. You can say no. Saying no does not change your supports, your placement, or how our team treats you.
How we handle information about you more generally is set out in our privacy and confidentiality policy, which is also available in Easy Read. Our website privacy policy covers the forms and analytics on this site.
If you think a rule has been broken
Tell any MMT Care worker, call 1300 066 822, or use the feedback form — you can do that without giving your name. You can also go straight to an independent body; our right to raise a concern page lists them. Nobody is treated badly for raising a concern in good faith.
Want to read the whole thing?
The full policy sets out the responsibilities of management and of workers, the kinds of AI tools it covers, and how AI risks are managed.