Safety, before you press start

MURMURLINE connects you with a random, unverified stranger. That's the whole idea — and it's worth a minute of context before your first chat.

You can't verify who's on the other end

There's no ID check, no account history, no profile. Your match could be anyone, anywhere, telling you anything. Treat it the way you'd treat any conversation with someone you've never met: don't take claims about identity, age, or location at face value, and don't let a stranger's urgency or friendliness rush you into sharing more than you intend to.

Practical tips

• Don't share identifying information you're not comfortable a stranger having — full name, address, workplace/school, phone number, social media handles, or financial details.
• Keep the conversation on MURMURLINE at first. A quick push to move to another app or platform is a common pattern worth being cautious about.
• You can leave at any time, no explanation needed. Hang up or skip — there's no cost to walking away from a conversation that feels off.
• If something feels wrong, trust that. You don't need a policy violation to justify hanging up.
• Never send money or gift cards to someone you met in a random chat, no matter the story.
• Report illegal content or conduct to local law enforcement in addition to using our Report button — we can act on your connection, but we can't investigate crimes.

How reporting works here

Press Report during a chat, or just hang up. Because messages are end-to-end encrypted (see how encryption works), our server cannot read chat content — so moderation is built without needing to:

• Your own browser screens your outgoing messages against known abusive patterns before sending, and can auto-flag a policy violation without ever sending the text itself to our server.
• A report or an auto-detected violation puts the offending connection into a temporary, graduated timeout — 5 minutes the first time, 30 minutes the second, 24 hours for further repeats within the same window. Every timeout has a fixed, visible expiry; there's no silent, permanent, unappealable ban.
• None of this requires — or involves — a human or a server reading your messages.

What moderation can't do

Because we deliberately can't read message content, we can't proactively scan conversations for harm the way a platform that stores plaintext might. That trade-off is intentional: it's the same property that makes the encryption meaningful. It does mean reporting is the mechanism that matters — use it.

If you're in immediate danger

Contact local emergency services. MURMURLINE is a chat relay, not a safety monitoring service, and has no way to identify who a real-world threat is beyond an IP-level timeout.