Every ranking on AI Models Map is governed by an independent council of five specialists. No single person controls the scores. Every methodology change is reviewed, debated, and published openly.
Each role brings a distinct lens to ensure rankings are accurate, secure, and fair.
A structured rhythm of checks, reviews, and audits ensures nothing slips through the cracks.
From automated pipeline to published rankings, every change follows a clear path.
Automated scoring pipeline processes all signals hourly.
System generates diffs when scores change beyond thresholds.
Council members review flagged changes for accuracy and fairness.
Approved changes go live with full changelog documentation.
Disputed changes are held for investigation or rolled back.
At step 3, the council can approve (Publish), request more data (Hold), or revert the change (Rollback).
Highlights from the Independence Charter that governs council operations.