Safety & moderation
Anonymous messaging done well is one of the most positive social experiences on the internet. Done badly, it's a harassment machine. Secret Message takes the difference seriously. Every design and engineering decision we make starts with the question does this make our users safer?
How moderation works
Every message sent through Secret Messageruns through a three-stage moderation pipeline before it ever reaches the recipient's inbox. The whole thing runs on our own servers as pattern matching. No external AI service ever sees what people write to you.
- Personal-data detection. Regex matching for email addresses, phone numbers, SSNs, IBANs, and card-shaped digit strings. A match routes the message to Filtered. Nothing is redacted or rewritten; you still see the message in full if you open the folder.
- Profanity wordlist. A soft signal only. Swearing on its own never filters a message, because most anonymous messages containing a swear word are affectionate.
- Severe-content patterns. Threats of violence, self-harm bait, sexual violence, hate slurs, and sexual content involving minors, written out across English, Turkish, Arabic, Spanish, French, Filipino, Indonesian, and German.
Messages matching the sexual-content-involving-minors patterns are blocked outright: never delivered, never stored, never in any folder. Messages matching the other severe categories, and messages containing personal data, are routed to your Filtered folder, where you can review them at your discretion or delete them unread.
The honest limit, because it decides what you should expect: pattern matching only catches what somebody wrote down in advance. It misses misspellings, coded language, new slang, inside references that only hurt because of shared history, and any language not on the list above. Cruelty phrased politely sails straight through, because there is nothing in it to match. It also produces false positives, which is why Filtered is a holding area rather than a shredder. Any service claiming its filter catches everything is marketing at you, and that gap is the reason the controls below exist and stay free.
What you can do
- Block senders. Based on the network connection the message arrived on. They will not be told they are blocked. Honest limit: someone who switches from wifi to mobile data looks like a new sender.
- Mute keywords. Any message containing them is silently routed to Filtered.
- Report messages. Flag for human moderator review. We triage within 24 hours.
- Delete messages. Individually, at any time, from the inbox. Deletion is immediate and permanent.
- Delete the whole board. This removes the link, the inbox and every message in it. The honest limit: while a board exists, anyone holding the link can send to it. There is no temporary pause switch, so if you want the messages to stop, the answer today is to delete the board and create a new link later.
Anonymity has limits
Senders are anonymous to the recipient. Senders are not anonymous to Secret Message or to law enforcement. We retain sender metadata (IP, country, device type) for 180 days. If we receive a valid legal request (subpoena, court order, or emergency-disclosure request) and the threshold for compliance is met under applicable law, we will disclose that metadata. We have committed to documenting law-enforcement requests in an annual transparency report, described further down this page.
If you receive something harmful
You don't have to deal with this alone. If a message threatens your safety, consider taking these steps in order:
- Report it inside the inbox (tap ⋯ → Report).
- Block the sender so further messages from that device or IP go nowhere.
- If you feel unsafe, reach out to one of the resources below.
- Contact local law enforcement if you believe there is a credible threat. Email [email protected] and we will preserve metadata while authorities investigate.
Crisis & mental-health resources
- United States: 988 Suicide & Crisis Lifeline. Call or text 988.
- United Kingdom: Samaritans. Call 116 123.
- Canada: Talk Suicide Canada. Call 1-833-456-4566.
- Australia: Lifeline. Call 13 11 14.
- International: Find a hotline at findahelpline.com.
Our commitments
- We will never paywall safety features.
- We will never offer a "reveal who sent it" feature, no matter how much money it would make.
- We will publish an annual transparency report covering law- enforcement requests and moderation outcomes.