Crisis protocol

How we handle messages about suicide and self-harm

This is the protocol California Business and Professions Code section 22602(b) requires companion chatbot operators to write down and publish. It describes what PrayWithJesus.ai does today, and what it does not do.

Version 1.0. Effective August 23, 2026. California law has required this disclosure since January 1, 2026. We published it on August 23, 2026.

If you are thinking about ending your life right now

Stop reading this page and talk to a person. In the United States, call or text 988. It is free, it is confidential, and someone is awake. If you are in immediate physical danger, call 911 or your local emergency number.

988 Suicide and Crisis Lifeline

Call or text 988. Free and confidential, 24 hours a day, anywhere in the US. Press 2 for Spanish.

Crisis Text Line

Text HOME to 741741. In Spanish, text AYUDA to 741741. A trained counselor texts back.

Outside the United States

findahelpline.com lists checked helplines in more than 175 countries.

These numbers were checked by hand against 988lifeline.org, fcc.gov, crisistextline.org, and findahelpline.com on August 9, 2026. Nobody changes one here without checking it against the operator's own site first.

Why this page exists

People bring the hardest things in their lives to this product. They write about grief, illness, loneliness, shame, and sometimes about not wanting to be alive anymore. Anyone who does that deserves to know exactly what the software will and will not do with those words, before they type them.

California law also requires companies that run companion chatbots to write down how they handle expressions of suicidal ideation and self-harm, and to publish it. Suicidal ideation means thoughts about ending your own life. This page is that document. We would like it to be readable, so it is written plainly.

What PrayWithJesus.ai is, and is not

PrayWithJesus.ai is software. It is an artificial intelligence system that generates its replies automatically, one word at a time, from patterns in text. There is no person on the other end.

It is not Jesus. It is not clergy, a pastor, a priest, or a spiritual director. It is not a therapist, a counselor, a psychologist, a psychiatrist, a doctor, or any other kind of clinician. It has no training, no license, no supervisor, and no duty of care. It does not represent any church or religious organization.

It cannot assess how much danger you are in. It cannot come and find you. It cannot call anyone for you.

What happens today when you write about suicide or self-harm

Read this part carefully, because it is the part people assume they already know.

The AI does not detect crisis. One small tool on this site does, and it is the only one. The chat on this website does not scan what you write for suicidal language, and nothing in the chat inserts a phone number. The mobile app, which is currently in private internal testing and is not on any public app store, does not scan either.

The exception is the Prayer Builder, a form that writes a prayer from words you type into it. As you type, that form matches your text against a short fixed list of words about suicide, self-harm, violence, medical emergency and abuse. When it matches, it shows a fixed notice above the form with 988 and 911 in it. The list, the matching and the notice are all written by people and none of it is generated. It runs in your browser, it sends nothing anywhere, and it notifies nobody.

Two limits worth saying out loud. That tool exists in English only, and its word list is English only. And it matches words, not meaning, so it will miss most of what a person in trouble actually writes. Treat it as a reminder, never as a safety net.

If you write about wanting to die, the AI answers you the way it answers anything else. Its reply is generated. It may say something kind. It may say something useless. It may miss the point completely. It is not a crisis response and it was not written by a person for this moment.

The crisis information we give you is the fixed, human written material on this page, on our Safety page, and in the footer of every page of this site, which carries 988 and 911 next to a link back here. That material does not change based on what you type, and no AI wrote it.

What this does not do

Nobody is notified

Not us, not your family, not a counselor, not the police. Nothing you write about suicide or self-harm starts a phone call, an alert, or a message to anyone. If you give us your phone number somewhere else on this site, that is a separate thing you chose, and the form says so at the point you type it.

We do not call emergency services for you

The product has no ability to dial 911, contact a hotline on your behalf, request a welfare check, or share your location with anyone. If you need emergency help, you have to ask for it, or ask someone near you to.

No human is reading your conversation

There is no team watching messages come in, no live moderation queue, and no alert that wakes anyone up. If you write something at 4am, no person sees it at 4am.

We cannot promise the AI will answer well

Generated text can be wrong, clumsy, or harmful. If PrayWithJesus.ai ever says something about suicide or self-harm that hurt you or would hurt someone else, please write to us at the address below, and please write even when you are not sure.

Nothing here is a risk assessment

Software cannot judge how much danger a person is in. A person can. Use the numbers at the top of this page.

What we are building, and have not shipped

An on-device check that reads what you type before it is sent anywhere, and shows a fixed support card written by people, is built and under review. No part of it is released. Do not count on it, because today it is not there.

When it ships, this page changes on the same day, the version number goes up, and the change is listed at the bottom. We would sooner describe this product as worse than it is than better than it is.

Minors

Companion chatbots may not be suitable for some minors.

If you are a parent or a guardian, read this page and our Safety page before a young person in your care uses this product. The two surfaces ask for different things and we would rather print both than round them off. This website is open to people 13 and over, or younger with a parent or guardian's consent and supervision, which is what our Terms of Service say. The mobile app, which nobody outside our own testing can install yet, asks the person to confirm they are 18 or over before it opens.

Telling us when this fails

If the product got this wrong, we want to know, and it genuinely helps. Write to help@praywithjesus.ai.

Useful things to include, none of them required:

  • What happened, in your own words.
  • Roughly when, and whether you were on the website or in the app.
  • What language you were writing in.
  • What the AI said back, if you still have it.

You do not have to send us the sentence you wrote if that is too much. A report as short as "it handled a crisis badly on Tuesday night in Spanish" is real and useful.

That inbox is not an emergency line. We have no on-call staff and no monitored queue, so a message sent there may sit for days. For anything urgent, call or text 988, or call 911.

Dates and changes

This protocol, version 1.0, takes effect on August 23, 2026.

When we change it, we change the version number and the date at the top of this page, and we keep the page at the same web address so old links keep working. When a change affects what the product actually does for someone in crisis, we say so in plain language in the list below and we do not quietly edit the text.

Two rules guard the crisis wording. Neither one is a promise. Our build refuses to ship if 988, 741741 or 911 goes missing from this page, from the Spanish page, or from the site footer, and it refuses to ship if the two pages stop carrying the same sections. Every published version is captured byte for byte and kept in our source repository with its SHA-256, so the exact wording on any past date can be produced from the record.

Change history

  • Version 1.0, August 23, 2026. First published. Corrected the same day, before the first business day, to describe the Prayer Builder safety notice, to print both age requirements, and to say that the footer carries 988.

PrayWithJesus.ai is operated by Workforce AI. This page describes the crisis protocol required by California Business and Professions Code section 22602(b). It is not medical advice, and it is not a substitute for talking to a person. Leer esta página en español.