Crisis Resources & Safety Protocol
Last updated: August 11, 2026
If you need help right now
If you are thinking about harming yourself, or you are worried about someone else, please reach out to a person. These services are free, confidential, and staffed around the clock.
988 Suicide & Crisis Lifeline (US)
Call or text 988
988lifeline.org - 24/7, also available via online chat.
Crisis Text Line
Text HOME to 741741 (US)
crisistextline.org - also serves Canada (686868) and the UK (85258).
Outside the US
Verified helplines in over 130 countries, searchable by country and topic.
Emergency services
911 (US) · 112 (EU) · 101 (Israel, Magen David Adom)
If someone's life is in immediate danger, call your local emergency number.
How Lifted responds to crisis language
This section is our published crisis protocol. It describes exactly what the app does, and does not do, when a conversation contains language suggesting suicidal ideation or self-harm.
What happens
- Automated detection. An automated check runs over messages as they are sent. It looks for language associated with suicidal ideation, self-harm, or immediate danger. It is a software check, not a clinical assessment, and it is not a diagnosis.
- The conversation stops steering. When the check triggers, Lifted stops the normal reflective flow. It will not continue prompting you to explore the topic, and it will not give advice about it.
- You see a supportive message and these resources. The app shows a short, non-judgemental message acknowledging what you said and surfaces the crisis resources listed above, including the number for your region where we can determine it.
- You stay in control. The app is not locked and your account is not suspended. You can dismiss the message and keep using Lifted, and the resources stay available from the app at any time.
What does not happen
- We do not contact the authorities, emergency services, or anyone else on your behalf. Detecting crisis language does not trigger a call to police, an ambulance, a hospital, a family member, or an emergency contact. No third party is notified.
- No one is monitoring your conversations in real time. There is no on-call team watching accounts. The check is automated, and it runs in the app and on our servers. It does not summon a human.
- Lifted does not perform a risk assessment. It cannot judge how much danger you are in, and it does not try to. That is what the trained people on the lines above are for.
- Detection is not perfect. It will miss things, and it will sometimes trigger on language that was not about self-harm at all. Never rely on Lifted to notice that you need help. If you need help, contact one of the services above directly.
Why we built it this way
Automatically escalating to emergency services on the basis of a text pattern can cause real harm: unwanted police contact, involuntary hospitalisation, and a strong reason for people never to be honest in the app again. We would rather put a trained human one tap away and leave the decision with you. If you would prefer a product that escalates on your behalf, Lifted is not that product, and you should not rely on it as a safety net.
What Lifted is
Lifted is an AI companion for self-reflection and journaling. It is not a medical device, it does not provide medical or mental health care, it does not diagnose or treat any condition, and it is not a substitute for a licensed professional. If you ask whether you are talking to a human or a licensed professional, it will tell you plainly that you are not. See our Terms of Service and Privacy Policy for more.
Reporting a problem
If Lifted responded to a sensitive message in a way that was harmful, dismissive, or wrong, please tell us. You can report AI output from inside the app, or email [email protected]. We read every report of this kind. Please note that this inbox is not monitored around the clock and is not a crisis line. For urgent help, use the resources at the top of this page.