Marketing Letters Маркетинг

Does AI reduce the reputation premium in knowledge diffusion?

Wang, Weijia
Springer Nature · 2026-09-27 · DOI 10.1007/s11002-026-09841-6

Аннотация

Merton ( 1968 ) showed that recognition accrues disproportionately to already-eminent producers of knowledge; marketing explains such patterns as signaling, with institutional affiliation acting as brand equity under quality uncertainty. Jo and Li ( 2026 ) report that AI-style writing attenuates this premium in SSRN working papers. This paper tests whether that claim generalizes to peer-reviewed venues and to algorithmic evaluators. In 476 Chinese business and management journal articles, institutional prestige predicts both downloads and citations, with elite-affiliated articles receiving 68% more downloads. AI-style writing independently predicts diffusion, but it does not moderate the prestige premium, and that null is precise rather than underpowered. In 546 SSRN working papers, the premium is small and only marginally significant. A third study asks whether machine readers reproduce it: four models from two independently developed families rated 100 of the articles with the affiliation shown or hidden, text held constant. Revealing an elite affiliation raises ratings relative to a non-elite one by 0.345 rating points; all four models are positive, and the effect is almost entirely a penalty to non-elite affiliations rather than a bonus to elite ones. The Matthew effect replicates; its strength is venue-dependent, and neither AI-style prose nor AI readers level it.

Читать у издателя →