A public, citable record of first-instance events in AI accountability. The moments that set the terms for what comes next. It exists because no one else was keeping the list.
The Register is a reference instrument, not a publication. Entries do not argue. They record. Analysis belongs in Vordan's Gap Alerts, Accountability Reports, VEPA assessments, or Architecture Reviews. The Register provides the evidentiary foundation those formats draw on.
Three conditions must all be met: Verification — at least one primary source must be identifiable and cited. First-Instance Specificity — the claim must be falsifiable; general category claims do not qualify. Accountability Relevance — the event must implicate assignment or avoidance of accountability for an AI system, or a structural condition affecting accountability at scale.
The Register does not claim comprehensiveness. It claims accuracy for what it includes. Challenges are evaluated within 30 days. Original entry text is never altered — corrections are dated and appended. Submit to hello@vordan.co.
Confirmed — Standard met; no substantiated challenge on record.
Disputed — A challenge has been received or found partially substantiated.
Superseded — A prior instance has been identified; entry remains as a research record.
No entries logged yet. First entries pending.