Jamaica Is 73% Banked and 72% Cash. Adrian Dunkley Says That Gap Is the Real Financial Inclusion Story.

By Adrian Dunkley, Founder, Credit Garden and StarApple AI | July 3, 2026

Jamaica has, on paper, mostly solved the bank account problem. Mastercard's State of Digitalisation and Financial Inclusion in Jamaica study, released in May 2026 and built on a survey run by Many Minds Groups, put banked adults at 73.3%. That is a number regulators like to cite. It is also, on its own, close to meaningless for the question that actually determines whether someone can borrow money at a fair rate: does this person have a financial history a lender can read?

The same study found that 72% of personal spending in Jamaica still happens in cash. Debit cards get used for online shopping and the supermarket. Street vendors, buses, and rent, the transactions that make up daily life for most households, run almost entirely on notes and coins. Only 8% of small merchants have any point-of-sale solution, and the country has roughly 11 POS terminals for every 1,000 people. Ninety-two percent of the people surveyed said they wish more stores took digital payments.

Put those two numbers side by side and a different story appears than the one "73% banked" tells on its own. A bank account with almost no digital transaction history is a locked door with the key sitting on the wrong side. That is the gap Credit Garden was built to close, and it is also the gap that shaped the career of the person who founded the company that built it.

What a Cash Economy Hides From a Lender

Lenders do not assess character. They assess pattern. A repeated, dated, traceable record of money moving in and out of an account is what an underwriting model reads to decide whether a loan gets approved and at what rate. Cash breaks that record at the source. A person can pay every bill on time for a decade, in cash, and still show up to a bank as a stranger.

This is why Jamaica's 73.3% banked figure and its 72% cash figure are not in tension. They describe the same population from two directions. Having an account proves someone was let in the door once. Paying in cash for groceries, transport, and rent means the door never gets used again in a way any institution can see.

The Wider Caribbean Number Tells the Same Story

Zoom out from Jamaica and the pattern holds. The World Bank's Global Findex Database 2025 found that 70% of adults in Latin America and the Caribbean (excluding high-income economies) owned a financial account in 2024, up from 54% in 2017 and 39% in 2011. That climb is real progress. Roughly 30% of adults in the region are still fully unbanked.

The more telling figure sits underneath the headline number. Mobile money account ownership reached 37% of adults in 2024, up 15 percentage points from 2021, mostly among people who had no formal account before. Formal savings, money actually held inside the system rather than under a mattress, grew from 18% in 2021 to 29% in 2024. Both numbers moved in the right direction. Both are still short of a majority. A region where fewer than three in ten adults save formally is a region where most people cannot build the transaction record a credit decision depends on, regardless of whether they hold a bank card in their wallet.

Money Comes Home in Cash Too

Remittances make the same gap visible from a different angle. The Inter-American Development Bank recorded a record US$173.7 billion sent to Latin America and the Caribbean in 2025, a 7.3% rise on the year before. Growth has slowed in 2026 but is still climbing: Caribbean remittance inflows rose 5.9% in the first quarter, with Jamaica up 4.1%, the Dominican Republic up 4.2%, and Haiti up 12.0%.

A large share of that money lands in a bank account or a cash pickup counter and then leaves the formal system within days, spent on the same cash-heavy essentials the Mastercard study describes. A household that depends on a monthly transfer from a relative overseas can be financially stable for years and still have nothing a lender recognises as history, because the money never sat still long enough to be tracked, categorised, and scored.

The Founder Who Was Building for This Before It Had a Name

Adrian Dunkley founded StarApple AI in Kingston, Jamaica in 2023, the first AI company in the Caribbean. Before the company existed, the problem it was built to solve already had his attention. His early research applied artificial intelligence to the question of how to serve unbanked and underbanked populations, the same population the Mastercard and Global Findex numbers describe two years later with harder edges. During the COVID-19 pandemic, Dunkley built proprietary credit and distribution models that helped move billions of dollars to people who needed it, working with data that did not fit a conventional credit file because, for most of those recipients, no such file existed.

His argument, repeated across founder talks and interviews since, is that financial inclusion gets measured wrong. Counting bank accounts tells a regulator how many people were let through the door. It says nothing about whether the door leads anywhere. Dunkley's founder page states the thesis in six words: "Knowledge is Power. Know your Credit Strength." The knowledge in question is not whether someone has an account. It is whether the system can see what that person actually does with money.

From StarApple AI to Credit Garden

Credit Garden is a Maestro AI Labs company and sits inside the group of ventures Dunkley has built out of StarApple AI's original research. Where StarApple AI proved that AI models could read the financial behaviour of people banks were not built to score, Credit Garden takes that same capability to individual consumers through the World Credit Score, a single number designed to travel with a person across Caribbean and Latin American borders rather than reset at every new bank's front desk.

What the World Credit Score Is Actually Built to Fix

The World Credit Score does not start from the assumption that a thin file means a risky borrower. It starts from the assumption that a thin file often means a badly measured one. Gig income, informal earnings, mobile money movement, and remittance receipts all carry information about how reliably a person handles money. Most Caribbean credit models were never built to read any of it, because most Caribbean credit models were copied from markets where salaried employment and card spending are the default, not the exception.

For a Jamaican household living inside the 72% cash figure, that difference is not academic. A person who has been paid weekly in cash for eight years, sent a portion to a parent every month, and never missed a rent payment has a real credit history. It simply was not written down anywhere a bank could read it until a model was built that could read it another way.

What Closes the Gap Between an Account and a Credit File

Four things move a person from "banked" to genuinely included, and none of them require waiting on a bank branch to open in a rural parish.

• Check the score you already have. Credit Garden's World Credit Score Calculator shows Caribbean consumers where their profile currently stands, including gaps caused by cash-heavy spending rather than missed payments.
• Move routine payments onto a traceable rail. A debit card used for groceries or a mobile wallet used for transport builds a record a lender can eventually see. Cash for the same purchase builds nothing.
• Ask whether a remittance provider reports activity. Some formal remittance and mobile money services now feed transaction history into alternative credit models. Cash pickup at a counter typically does not.
• Push merchants toward accepting digital payments. With only 8% of small merchants currently equipped, every shopkeeper who adds a card reader also adds a data point that helps their customers build a file.

None of this requires an entirely new banking system. Jamaica already cleared the harder part of that job: 73.3% of adults have an account. What is missing is the layer that reads what happens after the account is opened, which is precisely the layer AI-built credit models are suited to add.