How we use AI
A standards body that examines how others govern artificial intelligence should be examined the same way. This page states where AI works inside Quiet AI, what humans decide, and what is never automated.
Where AI works
The Quiet AI diagnostic engine performs the analysis behind every review. It gathers evidence, drafts findings, scores each criterion against QIA-1:2026, and checks every citation against its source across three passes: drafted, challenged, confirmed. AI tooling also assists in producing our published films and cards and in drafting correspondence.
What humans decide
Every flagged finding, every material omission, and every result is reviewed by the Quiet AI Standards Committee before a file is issued, and every issued file is signed by the Committee. Scoring is human-reviewed by design: no score, certificate, suspension, revocation or correction is issued by a fully automated decision.
What is never automated
- Issuing a result, at any review level.
- Suspending or revoking a certificate.
- Deciding a correction to an issued file.
- Publishing any claim about a named organization.
Beyond reviews, no Quiet AI system scrapes or automates third-party platforms. Our publishing runs through official interfaces only.
If you question a result
Write to apply@quietaistandard.org. Verified factual corrections are made within ten business days and recorded in the file’s verification record, as the methodology provides.
Why this page exists
The audience we serve is professionally sceptical, and the first question a sceptic should ask an AI-governance reviewer is how it governs its own. The honest answer: openly, with the engine doing the analysis and people carrying the accountability. We hold ourselves to the standard’s own pillars, and this page is part of that evidence.