The Source Register is the institution's list of the sources AI systems weight in their commercial answers, classified by tier and reviewed quarterly. Top 100 sources by weighted citation share, plus twelve vertical sub-registers — published in full each year as part of the Annual.
The Source Register is the dataset behind the Citation Source Quality signal in the AI Discovery Score. It is reviewed quarterly by the Methodology Council; admissions, removals, and weight changes are recorded with reasons.
Citation source quality cannot be assessed without a published source classification. The Source Register is the institution's published, examinable, amendable answer — the document any party may inspect to understand which sources count, why, and at what weight.
Sources are classified into three tiers, distinguished by editorial standard, independence, and durability of editorial governance.
Recognised professional bodies, accredited registers, university and government domains, established editorial publications. Independent editorial governance; durable across regulatory change.
Established trade press, recognised industry-specific media, credentialed directory and aggregator listings. Editorial governance present; weighting calibrated below Tier I.
General directories with editorial standards; comparison platforms with disclosed editorial methodology; local press for regional verticals. Used cautiously and only at appropriate weighting.
The Source Register is governed by five rules, each ratified by the Methodology Council and recorded in the Charter.
The full Source Register, including the top 100 sources and twelve vertical sub-registers, is published in the State of AI Discovery 2026.
Read the Annual →