Every entry on this site traces to a document a government body published. Here is how the record is made, and what it deliberately will not do.
Every status, figure, date, and coverage statement here is produced by a fixed software engine that reads archived primary documents and applies the same rules to every company. It is not AI-written opinion and not editorial judgment. The engine reports what a government document says, quotes a company’s own words where relevant, keeps proven findings separate from unproven allegations, and where the public record is silent it says exactly that. It ranks nothing, scores nothing, and draws no conclusions. Because the same rules run on every company from the same documents, the result is reproducible.
No entry appears without an archived primary document behind it: a court judgment, a consent decree, or an attorney-general filing, captured, hashed with SHA-256, and dated. The blue stamp on an entry means that document is on file.
We describe government actions in the government’s own words and quote a company’s marketing in the company’s own words. We never write the sentence that connects them. Findings, settlements, and pending allegations are kept in separate sections, because a pending allegation is not a finding.
When a search of government sources turns up no action against a company, we say exactly that, with the date and the sources checked. The absence of a located action is not a finding about the company; it is a statement about our search, and nothing more.
If a document is superseded or we get something wrong, the correction is made in the open and dated. Any company named here may submit a response, and we archive and attach it to the record it addresses. To report an error or submit a response, write to corrections@theexitrecord.com. Records current as of 2026-07-07.