Nurture-Chat - Your Parenting Partner From Day One

If you or your baby need help right now, don’t wait for an app. Call your emergency number: 999 in the UK, 911 in the US, 000 in Australia, 112 in the EU.

If you’re having thoughts of suicide or self-harm: in the UK, call Samaritans free on 116 123, any time. In the US, call or text 988. In Australia, call Lifeline on 13 11 14. Elsewhere, your local emergency number can put you through to a crisis line.

1. You’re talking to an AI

Nurture’s assistant is artificial intelligence. It isn’t a person, and it isn’t a doctor, nurse, midwife, health visitor or therapist — it never claims to be one, and if you ask whether you’re talking to a human, it tells you it’s an AI. It gives general information and support, not medical advice, and it can be wrong. See our Terms (section 4).

2. How it notices when someone might need help

Every message you send the assistant is checked automatically, as it arrives, for signs that someone may need urgent help, including:

signs of an emergency with your baby, such as trouble breathing or a seizure;
signs of an emergency for you after the birth, such as heavy bleeding, chest pain or a severe headache with changes to your vision;
thoughts of suicide or self-harm;
intrusive thoughts and feeling unable to cope — which are very common after having a baby, so we respond with reassurance first, then support.

This is automatic, and it is there to help — it isn’t a monitoring service. It can’t catch every situation, and it only sees what you write to the assistant.

3. What happens next

The reply puts safety first. For an emergency, the assistant opens by calmly telling you to call your emergency number now. For thoughts of suicide or self-harm, it also gives you a crisis line. For anything that needs a professional but isn’t an emergency, it suggests your GP, health visitor, midwife or NHS 111 — calmly, without alarm.
In the community, posts, replies and private messages are checked the same way (with OpenAI’s moderation as well). If one suggests the writer may be in crisis, they are shown a card with the right numbers straight away, and the post goes to the front of our team’s review queue.
It never helps with harm. The assistant is instructed never to encourage, describe or help with suicide or self-harm, in any conversation.

4. What we don’t do

We don’t contact emergency services, your family or anyone else on your behalf, and no one is watching in real time. If you need help, please reach out yourself, using the numbers above.
We don’t read your chats to watch over you. To check that these safety features work, we record only that a safety check was triggered, when, and what kind of concern it was — never your words (see our Privacy Policy, section 2).

5. Tell us when it gets something wrong

Under every answer there’s a Report link. Use it if an answer is wrong, unsafe or unkind — it comes straight to our team with the question and the answer, and we use it to make the assistant better. You can also email us at any time.

6. Keeping this up to date

We review this protocol, and how the assistant handles difficult moments, regularly and whenever we change the assistant. Questions: nurturechatteam@gmail.com.