Skip to content
Build n Bloom
NDIS
IndustriesAgentsResourcesBlogAbout
AI COMPLIANCECHATGPT & PROGRESS NOTES

Can your team use ChatGPT for progress notes?

The short answer is no — and most NDIS providers don't know their staff are already doing it.

It's 9pm. A support worker has six progress notes to write and twenty minutes before they want to be asleep. So they open ChatGPT, paste in what happened on shift — the participant's name, what set off the incident, how the behaviour presented — and ask it to write the notes up properly.

It comes back clean. They paste it into the system. Done in four minutes.

That worker just disclosed a participant's health information to an overseas company that, on a free or Plus account, can use it to train its models. By default. OpenAI's own terms: personal Free, Go, Plus and Pro accounts feed model training unless someone digs into settings and switches it off — and that opt-out only applies to what you type after you change it.

Progress notes are sensitive information under the Privacy Act. Disability, behaviour, health, all of it. Putting them into a consumer AI tool is a disclosure to an overseas recipient you no longer control — an APP 8 breach, possibly an APP 6 one, and the provider, not the worker, is the one the Commission holds accountable.

Here's the part that matters: the worker wasn't reckless. They were unsupported. Nobody gave them a compliant way to do the thing they obviously needed to do faster.

THE FIX

What is the fix for staff using ChatGPT for progress notes?

So the fix isn't a ban everyone ignores. It's two moves:

1

Give them a business-grade tool that doesn't train on what's entered — enterprise and team products don't, the consumer apps do. Run it on Australian infrastructure and the cross-border problem disappears as well.

2

Put one paragraph in the acceptable-use policy: no participant information goes into ChatGPT, the consumer Claude or Gemini apps, or any tool the organisation hasn't approved. Then make the approved tool genuinely easier than the banned one — or the policy is theatre.

The providers who get burned here aren't the ones whose staff are careless. They're the ones who left a gap and hoped.