Across the institution's first year of measurement, one finding emerged with such consistency that we treat it as a near-law: trust signal — not mention frequency — is the durable driver of how AI systems decide who to surface. This chapter sets out the evidence and the consequences.
This chapter sits at the centre of the inaugural State of AI Discovery 2026. The full Annual, with the methodology in support, is available on the Annual page.
Across 8,400 prompt-set responses spanning twelve indexed verticals, subjects with high mention frequency but low citation source quality were surfaced 38% less often than subjects with the inverse profile. The market spent the prior decade optimising for the wrong number.
The chapter's five findings, in the order they appear in the Annual. Each is supported by a methodology section with full prompt-set listing.
Subjects whose authority distance closed by ≥ 1 hop in 2025 retained 84% of their gain into Q1 2026. Subjects whose mention frequency rose without authority work retained 31%.
Across all five major systems, content volume from low-Tier sources is now demonstrably reranked downward. The 2024 effect, where volume produced visibility, is no longer observed in 2026.
One Tier I editorial citation produces visibility lift equivalent to 22 Tier III directory entries. The institution recommends ratios accordingly.
Pages without named, credentialed authorship are cited 47% less frequently than the same content with attribution. Across professional services the gap reaches 71%.
The gap between high-authority and low-authority subjects in the same vertical widened across all four observed quarters of 2025. The institution's forecast: it widens further in 2026.
Five operational shifts the institution recommends in light of the Authority Premium. Each is treated in detail in the Annual.
The Authority Premium is one of twelve chapters in the State of AI Discovery 2026. The Annual is available now.
Read the Annual →