What happens to your report after you press send
Trusting a website with an account of something that happened to you is a real decision. Here is exactly what the machinery does with it.
1. Nothing identifying is asked for
There is no account, no login, no email field and no phone field. The form asks about the office, the service, the amount and what happened — not about you.
2. Automatic redaction runs before a human sees it
The narrative is passed through a deterministic redaction pass that removes phone numbers, email addresses, ID numbers and vehicle numbers. The rules are versioned, so every report carries a record of which ruleset cleaned it.
3. A human reviews it
A moderator checks three things: that no personal detail survived, that the report is about a bribe demanded or paid for a public service, and that the structured fields match the account. Reports naming an identifiable individual are not published — the record is about offices, not people.
4. It is published as an anonymised record
What appears is the department, the city and state, the amounts, the outcome and the cleaned narrative, under a reference code. That code is how you can point to your own report without being connected to it.
5. Small slices are held back
Aggregates suppress any group thinner than a fixed floor, and that floor lives in the database view itself rather than in a query someone might forget. A slice too thin to be anonymous is never shown, no matter which page asks for it.
What is deliberately impossible
- No IP address is stored — only a salted one-way hash, used to slow spam.
- The published views cannot reach the raw drafts, structurally, so a read cannot return personal data.
- There is no way for us to contact you, which also means there is no contact list to lose or hand over.
More detail is on the FAQ, and the reasoning is in why anonymous reporting matters.
One minute, no account, in your language.
Report a bribe