Solacse
Safety & Crisis Protocols
Solacse is an AI companion for emotional support — not a human, not a therapist, and not a licensed medical provider. This page describes, in plain language, how we try to keep users safe and what we do when someone is in crisis. It is published in accordance with California SB 243 (effective January 1, 2026) and comparable disclosure requirements in other jurisdictions.
1. What Solacse is (and isn't)
Solacse is an artificially generated companion chatbot. It cannot diagnose, treat, or prescribe. It is not a replacement for a therapist, doctor, or crisis counselor and is not covered by medical confidentiality or legal privilege. Every response is generated by an AI model and may be inaccurate.
2. AI identity disclosure
A persistent banner and footer notice inform every visitor that they are interacting with AI. If a user ever asks whether they are speaking to a human, the assistant is instructed to say plainly that it is an AI.
3. Crisis cue detection
Each user message is screened for signals of self-harm, suicide, abuse, or imminent danger. When such cues are detected, Solacse:
- Interrupts the ordinary chat flow and surfaces a crisis card.
- Displays localized emergency numbers based on the user's region.
- Encourages the user to reach a human — a trusted person or a crisis line.
- Never provides means, methods, or content that could increase risk.
4. Localized emergency routing
We resolve the user's approximate region from network signals (never precise location) and prioritize the correct emergency resource — for example 988 in the United States, 116 123 (Samaritans) in the UK, 112 across the EU, and iCall/Vandrevala in India. A global directory is always shown as a fallback.
- United States988 Suicide & Crisis LifelineCall or text 98824/7, free, confidential
- United KingdomSamaritansCall 116 12324/7, free
- IndiaiCall / VandrevalaCall 9152987821Mon–Sat, 8am–10pm
- CanadaTalk Suicide CanadaCall 1-833-456-456624/7
- AustraliaLifeline AustraliaCall 13 11 1424/7
- GlobalFind a Helplinefindahelpline.comLocal lines worldwide
5. Professional referral
When distress is sustained across a conversation, when a user explicitly asks for human help, or at any time via the always-visible referral button, Solacse invites the user to connect with a vetted human professional through our directory. Professionals are approved only after credential and identity review.
6. Minor safety (age gate)
Solacse is intended for users aged 18 and above. During signup we require explicit confirmation of age. Content generation is tuned to avoid material that could encourage unsafe behaviors or induce extreme emotions, consistent with China's interim measures on anthropomorphic AI and comparable minor-safety expectations globally.
7. Ongoing psychological safety review
A scheduled backend audit samples anonymized assistant responses and evaluates them against safety baselines — crisis handling, refusal quality, tone, and adherence to the AI-identity disclosure. Findings feed back into prompt and policy updates. No personally identifying content is included in the audit log.
8. Data retention and legal discoverability
Chats are stored so we can pick up where you left off. Please be aware that these records can be subject to legal discovery (for example in civil disputes). See the privacy policy for details and deletion options.
9. Contact
To report a safety concern, please use our report form. For emergencies, contact your local emergency number immediately.