Generative AI Adoption for Image and Video Creation by Creators and Small Businesses in Emerging Markets: An Evidence Map and Early Platform Usage Study

Abstract

Creators, freelancers, and small businesses in emerging markets are among the earliest and most intensive users of generative image and video tools, yet the academic evidence on their adoption remains fragmented across disciplines. This article contributes a two-part evidence base. The first part is a verified corpus of 850 unique academic works on generative AI adoption, assembled through 190 audited queries across OpenAlex, Crossref, and arXiv; every record is deduplicated, 838 carry a resolvable DOI, and 762 are confirmed by two or more independent bibliographic sources. Coverage analysis of the corpus quantifies a specific gap: only nine works simultaneously address emerging markets, image or video generation, and creator or small-business populations, and the intersection of video generation with small and medium-sized enterprises remains empty. The second part supplies behavioral evidence from a 45-day production window of Geramaker, a Brazilian pay-per-use platform for AI image, video, narration, and music generation, covering 2,605 accounts, 7,005 model calls, and 62 paid transactions. Adoption proves near-instant under free entry (median of one minute from signup to first output), purchase decisions compress to minutes on local instant-payment rails, and image-to-video animation of existing photographs dominates video demand. Retention, rather than activation, emerges as the binding constraint: 96% of users generate on a single calendar day. The article maps the observed drivers and inhibitors onto established adoption constructs and derives a research agenda for the under-studied intersection the corpus reveals.