Safety
What Delia is, and is not, in a hard moment.
She is a private notebook, not a crisis service. Here is exactly what happens if you tell her you're in real danger, and what never does.
If you tell her you're in danger.
The moment a message shows real distress, Delia stops. She does not try to talk you through it, does not improvise comfort, and does not generate a reply at all. Instead she gives you the same fixed response every time, word for word: 988, the Crisis Text Line, 911, and a plain statement that she's not the right kind of help for what you're facing right now, but these are.
That response is checked, mechanically, against the exact wording it is supposed to be before it ever reaches you. If it doesn't match byte for byte, it doesn't ship — a person reviews it instead. A hard moment is the wrong place for a chatbot to get creative, so this is the one place in Delia where nothing is generated at all.
Resources that are always one tap away.
Not gated behind a conversation. Listed plainly in Settings, any time.
- 988 Suicide & Crisis Lifeline — call or text 988, in the US
- Crisis Text Line — text HOME to 741741
- SAMHSA National Helpline — for substance use and mental health, 1-800-662-4357
- Eldercare Locator — for older adults and their families, 1-800-677-1116
- AARP Friendly Voice — a free program pairing older adults with a caring volunteer to talk to
- National Domestic Violence Hotline — 1-800-799-7233
- The Trevor Project — for LGBTQ young people in crisis, 1-866-488-7386
Who can use Delia.
Adults, 18 and up, checked at signup. She launches in the United States, on iPhone first. She is not available in Illinois, a deliberate choice under that state's restrictions on AI systems that could be mistaken for therapy.
What she will never do.
- She never diagnoses, treats, or claims to offer therapy.
- She never keeps a hidden distress flag or a running file on your emotional state.
- There is no dashboard where anyone watches your words for warning signs.
- There is no separate "crisis mode" personality — she does not perform calm, she simply speaks plainly and points you to real help.
- She never infers a welfare check or contacts anyone on your behalf without you asking her to.
Independent review.
The safety protocol itself — the exact resources, the exact wording, the age gate, the state restriction — is reviewed and signed off by an outside clinical advisor before launch, and again every year after.
This page describes the same protocol as ourprivacy commitmentand the straight answers page — nothing here is softened for marketing.