Crisis protocol

Updated October 30, 2026. Atiq Litera is for adults 18 and over. You are talking to an AI, not a human. It is not a crisis service.

What the screen checks

Before a chat message reaches a voice, a fixed screen checks it for signs of suicide, self-harm, harm to others, abuse and intrusive thoughts. The AI voices are also instructed to flag immediate danger or self-harm.

What happens when danger is detected

When the screen detects these signs, the message receives fixed safety support instead of a voice answer. When a voice flags a crisis, its generated answer is replaced by fixed safety support outside the role. These messages use no answer credit. The support directs people to real human help, including local emergency services, 988 in the US and findahelpline.com. Each time the site gives safety support for a message, one is added to a daily count; the count keeps no message text and nothing that identifies you.

If you or anyone else is in immediate danger, contact local emergency services now. In the US, call 988 or text 988 (Suicide & Crisis Lifeline) any time. Elsewhere, find a free helpline in your country at findahelpline.com.

If risk continues

Each new chat message goes through the same screen. After support for danger to your own life is shown, the voices continue to receive crisis guidance for the rest of the visit: across voice changes and reloads, and into your account when you sign in on this device within six hours. Support shown while you are signed in with saved history turned on is also saved with your recent history, so your next answers on other devices receive this guidance too. Signing out or switching to another account ends it on this page. Support for abuse, harm to others or unwanted intrusive thoughts keeps that guidance available for the next two messages.

Contact

Contact the operator at support@atiqlitera.com. If you need urgent help, use emergency services or a helpline.

Read the terms · Read the privacy notice · Return to Atiq Litera