✦ COMPOSURE

Safety and Crisis Protocol

Last updated: August 17, 2026
If you need help right now

United States: call or text 988, the Suicide and Crisis Lifeline, any time.

UK and Ireland: call 116 123 to reach Samaritans, free, any time.

Anywhere else: findahelpline.com lists services by country.

If you are in immediate danger, call your local emergency number.

What Composure is

Composure is a practice tool. You rehearse conversations with fictional AI characters and get feedback on how you came across. It is not therapy, counseling, or mental or behavioral health treatment, it is not a substitute for any of them, and nothing in it is delivered or supervised by a licensed professional.

No person reads your sessions, and nobody is monitoring them in real time. Composure cannot notice that you are in trouble the way a friend or a clinician could, and it cannot send help.

How we handle a crisis moment

Our practice characters are built to stay in character. That is right for rehearsing a hard conversation and badly wrong if someone says something that signals real distress, so we never let a character be the thing that answers in that moment.

Every message you speak or type is screened before it reaches any AI model. If it contains signals of suicidal thinking, self-harm, or intent to harm someone else, then:

The screening runs on our own servers using pattern matching, not an AI model. Everyday figures of speech such as "this interview is killing me" are deliberately excluded so the screen appears when it matters and not otherwise.

What we do not do

Age and minors

Composure is for people 13 and older. If you are under 18, you should use it with the consent and involvement of a parent or guardian. Our characters are instructed to refuse sexual content, and repeated attempts to elicit it can limit access to the app.

Reporting a problem

If Composure ever responds to something serious in a way that feels wrong, please tell us at composure.help@gmail.com. Include roughly when it happened and what was said if you are comfortable doing so. Safety reports are read first.

Why this page exists

California Senate Bill 243, in force since January 1, 2026, requires operators of companion chatbots to maintain a protocol for responding to expressions of suicidal ideation, self-harm, or suicide, to refer users to crisis services, and to publish that protocol. This page is that protocol. Comparable requirements exist in New York and Nebraska.