AI Visibility of Latin American Fintechs: A Share of Voice and Technical Readiness Study Across Five Countries

This paper presents the first systematic measurement of AI generative engine visibility for Latin American fintech companies. Using a corpus of 26 B2B queries executed in Claude Sonnet 4.6 across six product categories and five Spanish-speaking markets — Argentina, Mexico, Colombia, Chile, and Peru — we document the AI Share of Voice (SoV) of 28 companies and audit their technical readiness to be cited using an adapted version of the Entity Visibility Score (EVS) framework. Key findings reveal near-total geographic compartmentalization of recommendations (only Mercado Pago achieves top-5 presence in all five countries), a paradox whereby the ecosystem's most-cited company actively blocks AI crawlers in its robots.txt, and a systematic advantage of B2B infrastructure providers (Belvo, dLocal, MetaMap) over consumer-facing neobanks in cross-country AI visibility. No company in the sample has implemented llms.txt, representing an unaddressed sector-wide gap. We also document the implications of Anthropic's February 2026 bot architecture update — which separates ClaudeBot, Claude-User, and Claude-SearchBot — for fintech companies' crawling policies. Practical recommendations are provided for multi-country operators, B2B infrastructure providers, and marketing teams.