What this site records, what it does not, and what happens to anything you send us.
No public accounts, no advertising, and no third-party tracking. The only sign-ups are two optional email lists: the weekly email and email alerts. A first-party counter records only approximate daily visitors and page views; personal information is otherwise held only when you submit a correction request, send a message through the contact form, or sign up for an email list.
Operational logs record request time, method, path, and status. For API requests we also log the user-agent, referrer, query and a keyed hash of the network address, kept for about 8 days to spot abuse and heavy automated use. We never store the raw address. Transient in-memory rate-limit counters use the network address. No profiles are built, and we don't use these logs for advertising or to identify readers.
Each page load and in-site navigation sends an empty event to this site's own server. It contains no page path, title, referrer, campaign data, screen details, or client identifier. A secret-keyed, day-specific transformation of the network address estimates daily unique visitors without storing the raw address or creating a cross-day identifier. The day-specific codes expire after two completed UTC days; only daily aggregate visitor and page-view totals remain.
No advertising or public-visitor analytics cookies. One functional cookie stores a sidebar preference; hosting infrastructure may set a technical routing cookie. A signed eight-hour security cookie is set only after the owner signs in to the private analytics page.
A correction request stores the request details, proposed value, justification, contact details, and source network address. Personal content is encrypted, visible to reviewers only, excluded from public pages, the API, and exports, and erased automatically after 12 months while the decision remains.
A contact message stores your optional name, reply address, topic, and message, encrypted and readable only by the site owner, who also receives an email copy through a transactional email service. A keyed one-way code derived from the network address, never the address itself, limits abuse. Messages are deleted automatically after 12 months.
The weekly email and email alerts each need your email address; alerts also need the up to three companies or domains you choose to watch. Both are stored encrypted, used only to send what you asked for, and never shown on the site, returned by the public API, or shared. Emails are sent through a transactional email service.
Apart from the confirmation email, nothing is sent until you click its link. Confirmation links expire after 48 hours, sign-ups that are never confirmed are deleted after seven days, and a keyed one-way code derived from the network address, never the address itself, limits sign-up abuse.
Unsubscribing deletes your encrypted address and, for alerts, your watchlist; only a one-way fingerprint of the address and the unsubscribed status remain, so you are not emailed again unless you sign up again. Records of which alerts were sent exist only to avoid repeats and are deleted after 60 days or on unsubscribe; weekly-email delivery records hold no address. Lists checked with the vendor check are used once and never stored or logged.
The site republishes narrowed allegations: attacker prose, leaked-material links, contacts, and infrastructure are stripped, and rows that fail safety checks are quarantined. Affected organizations can use the claim correction form on any claim page.
No account data normally exists. Email-list subscribers can leave at any time with the unsubscribe link in every email, which deletes their address. Correction submitters can request early erasure, and the functional cookie can be blocked without losing access. Privacy questions go through the contact form; see also the terms of use.
Victim records are provider-reported allegations, not independently verified breaches. Published for research, defense, and educational use.