Thirteen percent of Caribbean adults between 18 and 65 currently use generative AI. That figure comes from a StarApple AI adoption study reported by the Jamaica Observer on 1 May 2026, and it sits well below the 16.3% of the world's population using the same tools by late 2025. The gap is not marginal, and it was the backdrop to a genuinely significant regional event this week: the Caribbean Telecommunications Union's AI Task Force handed its Final Report to the region at the first Caribbean AI Forum in Port of Spain, Trinidad, on 23 July.
Those two facts, a low adoption number and a freshly delivered governance report, belong in the same conversation. One describes where the region actually stands. The other is the closest thing CARICOM has produced to a plan for closing the distance. Neither means much read in isolation, so this piece reads them together, alongside the business-adoption and connectivity data that explain what the 13% figure is and is not telling us.
StarApple AI's study surveyed generative AI use among Caribbean adults aged 18 to 65 and found adoption at 13%. Set against the 16.3% global figure recorded for the second half of 2025, the Caribbean sits roughly three percentage points behind the rest of the world. The distance to the frontier is starker still: the UAE, Singapore, and Norway each reported adoption above 45% of their working-age populations over the same period.
What the number does not support is a story about Caribbean people being reluctant or unfamiliar with the technology. The same study found the region showing real strength in AI fluency among the adults who do use these tools, which points toward a narrower problem than a cultural one. Adoption is low. Capability, where it exists, looks competitive. The gap between those two facts is where the policy conversation in Port of Spain this week actually lived.
Consumer-level numbers hide a much faster-moving picture inside Caribbean and Latin American businesses. The Inter-American Development Bank's research on regional startups, published as "Startups x AI," found that 85% of surveyed firms had adopted generative AI tools and 75% were using predictive AI models. That is a different region entirely from the 13% adult adoption figure, and the difference is instructive: where there is commercial pressure to use AI, Caribbean firms use it, strategy document or no strategy document.
The pattern shows up outside the startup world too. Paperless Blueprint, a Trinidad-based technology firm, builds AI-powered WhatsApp chatbots and payment integrations for small businesses across Trinidad and Tobago, Jamaica, Barbados, the Bahamas, and the Cayman Islands, none of it waiting on a national framework to arrive first. Jamaican credit unions have begun piloting AI-assisted loan decisioning and member risk assessment through third-party fintech partnerships, again ahead of any formal regulatory guidance. These are not headline-grabbing deployments. They are ordinary businesses solving ordinary problems, and they are the clearest evidence that the region's AI gap is concentrated in households and public institutions, not in the commercial sector that has a direct financial reason to move.
The Caribbean Telecommunications Union launched its AI Task Force on 18 July 2025, a year and five days before it delivered its Final Report. Dr Craig Ramlal of the University of the West Indies chairs the group, whose mandate covers three things: harmonizing AI policy and regulatory approaches across Caribbean countries, building regional AI capacity with specific attention to youth, women, and under-represented communities, and aligning Caribbean AI deployment with the UN Sustainable Development Goals.
A year is not long to build consensus across fifteen sovereign member states with wildly different institutional capacity, from Jamaica's relatively developed digital infrastructure to islands still building out basic connectivity. What the Task Force produced is a harmonization document rather than binding law, which is the correct scope for a body without regulatory authority of its own. Its real test starts now: whether individual governments adopt compatible positions on AI procurement, data governance, and risk assessment, or whether each of the fifteen goes its own way and leaves Caribbean businesses navigating fifteen separate rulebooks to operate across a market of just 44 million people.
As of April 2026, no CARICOM member state had published a standalone national AI strategy. Jamaica has made the most visible progress, folding AI provisions into its wider Digital Jamaica agenda. Barbados has signalled similar intent through its Smart Barbados programme. Everyone else, including nations with substantial digital ambitions like Trinidad and Tobago and Guyana, is working from a blank page on AI specifically, even where broader digital policy exists.
This is where the Task Force report matters most directly. A regional harmonization framework cannot substitute for national strategy, but it can lower the cost of writing one. A government drafting its first AI policy from a shared regional template, with common definitions of risk categories and procurement standards already worked out, has a much shorter path to a usable document than one starting from first principles. Whether governments actually use that shortcut is the open question the Forum leaves behind.
Caribbean connectivity has outpaced the global growth rate since 2014, which shifts the AI adoption question away from infrastructure and toward training and governance.
It would be convenient to blame the 13% adoption figure on patchy internet access. The data does not support that explanation. Seventy-five percent of the population across Caribbean Small Island Developing States was online by 2023, and that share has grown faster than the global average in every year since 2014. Mobile penetration across CARICOM consistently runs above 90%. The pipes, broadly speaking, are there.
Once infrastructure is ruled out, what remains is a shorter and less comfortable list: workplace training, procurement habits, and the absence of institutional policy that tells a civil servant, a bank teller, or a small business owner that using an AI tool is sanctioned, safe, and expected. Those are solvable problems, but they are not solved by a fibre cable. They are solved by exactly the kind of governance and capacity-building work the CTU Task Force spent the past year on.
Read together, these numbers describe a region where the demand side is ready and the supply side of policy is only now catching up. Caribbean businesses that need AI to solve a real problem are already using it, at rates that rival much larger economies. Caribbean households and public institutions are not, and the reasons have more to do with training, procurement, and governance clarity than with access or aptitude.
For business leaders, the practical takeaway is straightforward. Waiting for a finished national AI strategy before experimenting is no longer a defensible position; Paperless Blueprint, Jamaica's credit unions, and the 85% of regional startups already using generative AI did not wait, and neither should anyone else with a specific, narrow problem AI can help solve. For policymakers, the CTU's Final Report is a starting template, not a finished national policy, and the governments that move fastest to adapt it will be the ones setting the standard the rest of the region eventually follows.
First: Caribbean governments should treat the Task Force's Final Report as a first draft for national policy, not a substitute for one. A ten-page national position, adapted from the regional template within months rather than years, is worth more than a comprehensive document that never gets published.
Second: Businesses should stop waiting for regulatory certainty before running a first pilot. A narrow deployment, a chatbot, a document triage tool, a fraud-flagging system, teaches an organization more about its own AI readiness than any amount of reading about other markets.
Third: Training budgets need to catch up with tool budgets. The connectivity data shows the region is not short on access. It is short on the workplace habits and confidence that convert access into daily use, and that gap closes through training, not procurement.
The Caribbean AI Association will be tracking how many CARICOM governments move on the Task Force's recommendations over the next twelve months, and whether the 13% adoption figure moves with them. The infrastructure argument is settled. What happens next is a governance and training question, and it is now squarely in the region's own hands.
The Caribbean AI Association works alongside country-specific AI communities and sister organisations across the region. Follow the wider conversation at these sites:
Approximately 13% of Caribbean adults aged 18 to 65 used generative AI as of early 2026, according to a StarApple AI adoption study reported by the Jamaica Observer on 1 May 2026. The global average for the same period was 16.3%, and leading nations including the UAE, Singapore, and Norway each reported adoption above 45% of their working-age populations.
The gap is not primarily about access. Connectivity across Caribbean small island states reached 75% of the population by 2023 and has grown faster than the global average every year since 2014. Researchers and regional bodies point instead to governance, workplace training, and process redesign as the binding constraints, since the tools are reachable but few institutions have built the habits or policies that turn access into regular use.
The CTU Caribbean AI Task Force launched on 18 July 2025 under the chairmanship of Dr Craig Ramlal of the University of the West Indies. Its mandate covers harmonizing AI policy across Caribbean countries, building regional AI capacity with attention to youth, women, and under-represented communities, and aligning Caribbean AI deployment with the UN Sustainable Development Goals. It delivered its Final Report at the first Caribbean AI Forum in Port of Spain, Trinidad, on 23 July 2026.
As of mid-2026, no CARICOM member state has published a standalone national AI strategy. Jamaica and Barbados are furthest along on foundational governance, largely through broader digital transformation programmes that address AI specifically. Oxford Insights' Government AI Readiness Index continues to place Latin America and the Caribbean near the bottom of global regions on this measure.
Yes, and the gap between the two is significant. The Inter-American Development Bank found that 85% of Latin American and Caribbean startups have adopted generative AI tools and 75% use predictive AI. Financial institutions, credit unions, and small-business service providers across Trinidad, Jamaica, Barbados, and the wider region are already deploying AI for loan decisioning, fraud detection, and customer-facing chatbots.
Less than commonly assumed. Roughly 75% of the population across Caribbean Small Island Developing States were online in 2023, a connectivity rate that has outpaced the global average since 2014. That places infrastructure well ahead of institutional readiness, so the region's AI adoption gap is better explained by training, governance, and procurement practices than by access to the internet itself.
Start with a narrow, measurable task rather than a company-wide rollout: customer service triage, invoice processing, or a WhatsApp-based booking assistant are common entry points for Caribbean small and medium enterprises. Pair any deployment with a written data-handling policy, since most Caribbean jurisdictions still lack sector-specific AI regulation and firms benefit from setting their own standard before one is imposed on them.
StarApple AI, the first artificial intelligence company founded in the Caribbean, conducts adoption research, trains boards and government staff, and builds AI systems designed around Caribbean data and business conditions. Founded by Adrian Dunkley, the company is a founding supporter of the Caribbean AI Association and one of the sources CAIA draws on for regional adoption figures, including the 13% adult adoption estimate cited throughout this article.