Responsible use of AI

What indi's AI does

• Organises what you share. Your chats, photos, documents and emails go into your child's timeline and records, so you don't have to sort them yourself.
• Writes down what you say. You can talk instead of typing, in chat or when you add a note, and indi turns your words into text. Our speech-to-text provider is contractually excluded from using your audio to improve its models, and doesn't keep it after transcription.
• Answers your questions. In chat, indi draws on what you've shared about your child.
• Helps you get ready for appointments. It puts together a summary from your child's records to take with you.
• Brings summaries to your child's clinicians. indi collates what you've shared into summaries your child's clinicians can read. What it means for your child's care is their call.

What it doesn't do

• It doesn't diagnose your child. It can talk through what you're noticing and what to raise with your GP, paediatrician or clinician. A diagnosis is their call.
• It doesn't make decisions about your child. You and your child's clinicians decide about their care, and funding decisions, such as an NDIS plan, are the funder's.
• It isn't a replacement for professional advice.

You stay in charge

• You can correct or delete anything indi has recorded about your child.
• When indi drafts a report for a clinician, the clinician reviews, edits and finalises it before your family can see it.
• Help improve indi is on unless you turn it off. While it's on, we may use your family's conversations and documents to work out how to make indi better. You can turn it off at any time in your privacy settings, and we stop using them from then on. Safety and quality checks are separate, and carry on either way. Here's how they work.

Your information and AI

• We don't use identifiable personal information to train AI models.
• The AI providers we use don't train their models on information sent through indi.
• Your information is encrypted when it's sent and when it's stored.

How we check our AI

• We routinely run automated checks on indi's responses, including responses to a parent in crisis, to assess their quality and safety.
• How we monitor our AI is set out in our internal AI Safety and Quality Monitoring Policy, owned by our Chief Technology Officer.

When a person reads a conversation

• A person on our team may read a response from indi, and what led to it. It's how we catch a response that's unsafe or wrong, and manage any risk of harm.
• What we may read: responses you mark as unhelpful or wrong, responses our automated safety and quality checks flag, and the messages around them in the same conversation.
• We read them as you entered them, without removing identifying details. The real words are what show us whether a response was safe and right.
• Only our own people do this inside indi, and only those who need to. What they find is kept with your child's record and deleted when the conversation is deleted.
• This carries on whether or not Help improve indi is on.
• When you ask us for help, our support team may see your information as you entered it, and only to help with what you've asked.

Questions?