ChatGPT's fastest-growing markets are Haiti, Tanzania, and Papua New Guinea, not Silicon Valley

ChatGPT's top per-capita rank movers are in developing economies, not Silicon Valley

ChatGPT adoption is accelerating fastest in developing economies, not the wealthy tech corridors where the industry concentrates its attention. Haiti and the Dominican Republic each jumped nine places in global per-capita usage rankings between Q4 2025 and Q1 2026, according to OpenAI's Q1 2026 Signals report. Japan rose eight places, Mexico and Tanzania gained six each, and Brazil, Costa Rica, Myanmar, and Papua New Guinea each moved up five 1.

The dataset covers consumer ChatGPT messages from July 2024 through March 2026 and excludes Enterprise, education, and Codex accounts 2. OpenAI ranks countries by messages sent per capita rather than total volume, meaning a smaller nation can climb if its residents send proportionally more messages, not necessarily because more people signed up. The ranking measures relative movement in position, not absolute market size 1.

Seven of ten fastest-rising countries sit outside the high-income tier

Mapping the ten fastest-rising countries to World Bank income groups reveals a structural pattern beneath the headline numbers:

• Lower-middle income: Haiti, Tanzania, Myanmar, Papua New Guinea
• Upper-middle income: Dominican Republic, Mexico, Brazil
• High income: Costa Rica, Japan, Austria

Seven of the ten are non-high-income economies 3. Costa Rica itself only graduated to the high-income tier in July 2025, having previously sat in the same upper-middle bracket as the Dominican Republic and Mexico 3.

This is not a new pattern. A National Bureau of Economic Research working paper by OpenAI's Economic Research team and Harvard economist David Deming, released September 2025, found that as of May 2025, adoption rates in the lowest-income countries were more than four times those in the highest-income countries 4. Haiti climbed from rank 91 to 82. Tanzania rose from 102 to 96. The Dominican Republic moved from 53 to 44 5.

Why a single tool compounds faster where infrastructure is thin

Where computing infrastructure is sparse, a conversational interface that bundles search, translation, writing, and calculation into one product carries more marginal value than it does in markets already saturated with specialized software. A user in Papua New Guinea gaining access to ChatGPT acquires capabilities that would otherwise require separate applications, paid subscriptions, and hardware. In wealthy markets, ChatGPT competes with established tools for the same tasks. In developing markets, it fills gaps that nothing else fills.

The demographic data tracks the geographic broadening. OpenAI reports that users with typically feminine names now account for more than half of ChatGPT users where gender can be inferred from names, having reached approximate parity in 2025 1. The NBER paper traced the split from 37% female in January 2024 to 52% in July 2025 4. Users over 35 gained share during Q1 2026, indicating adoption is spreading beyond early adopters 1. The fastest-growing work-related uses on consumer plans included visual design, health and medical documentation, operational planning, and proposal and grant preparation 1.

The data comes from one company measuring its own product

Every figure in the Signals report originates from OpenAI's own product data. The company has a commercial interest in demonstrating broad adoption, and the gender inference relies on name-based assumptions that may not work reliably across non-Western naming conventions. The per-capita ranking obscures absolute scale: India, the second-largest total user market by volume, does not appear on the fastest-rising list because the metric normalizes for population 1. Enterprise and education usage, which OpenAI excludes from this dataset, could tell a different story in wealthy markets where institutional purchasing concentrates.

For investors and builders focused on enterprise AI in wealthy markets, the Signals data points to a parallel audience forming habits elsewhere. The use cases climbing fastest are operational planning, grant preparation, and health documentation. These are not the chatbot novelty tasks that defined early ChatGPT coverage. They are work tasks performed on consumer accounts in countries where formal AI training and workplace policy have not arrived. Whether the companies building AI tools serve these markets or treat them as afterthoughts while competing for Fortune 500 contracts will shape who owns the next billion AI users.