Soulward Safety Protocol
Version of 27.09.2026.
Soulward is not an emergency service. If you are in danger, call 911 or 988 in the US, 112 in Europe, or your local emergency number.
How the assistant responds to a crisis
- Before a message reaches the AI, an automatic filter checks it for signs of suicidal thoughts, self-harm and overdose.
- If such signs are found, the message is not sent to the model. The user immediately gets a pre-written reply with emergency numbers and the in-app SOS button instead of an AI reply.
- The assistant is instructed never to describe methods of self-harm, doses or ways of using substances, or to encourage harm to self or others.
- The assistant discloses that it is AI at the start and whenever asked. It never presents itself as a doctor, psychologist or therapist.
- The app is only for people aged 18 and over.
Limitations
The automatic filter may miss a crisis or trigger by mistake. No human reads conversations in real time. In a dangerous situation do not wait for the assistant; call emergency services.
Contact
[email protected]