Trust & Safety

Our Approach to Safety

Plain language, no legal jargon — what we actually do, and what we don't.

How crisis detection works

Recover Sober uses an automated, code-based safety system — not an AI guessing in the moment — that scans messages in the AI Coach, live classroom chat, and Community posts for language associated with crisis or serious distress. This check runs before a message ever reaches the AI or gets published.

What happens when something is detected

If crisis language is detected, the system immediately blocks the AI response (or the public post) and surfaces emergency resources — the 988 Suicide & Crisis Lifeline and 911 for immediate danger — instead of generating a normal reply. For less urgent but still concerning language, a pre-written, carefully considered response is shown rather than a public post or an AI-generated reply.

What this is — and isn't

This detection is automated pattern-matching, not real-time human monitoring. No one is watching your conversations live. It's a safety net, not a substitute for professional care: Recover Sober is a digital recovery education and support platform, not a licensed treatment provider, and nothing here replaces therapy, medical care, or licensed addiction treatment.

Where we are today

We don't yet have a formal clinical advisory board reviewing our content and safety protocols. Adding licensed clinical oversight is an active priority as the platform grows, and we'd rather tell you that plainly than overstate where we are.

In an emergency

If you or someone you know is in crisis, please call or text 988 (US Suicide & Crisis Lifeline) or call 911. Don't wait on an app to respond.

Related

See our Privacy Policy for how safety-event data is handled, or our Terms of Service. Questions? Visit our Contact page.