StarApple AI | Adrian Dunkley | July 1, 2026

State of Caribbean AI: The H1 2026 Review

Six months of task force reports, national assessments, one 13 percent adoption number that undersells what is actually happening, and a Kingston-built AI platform that wants to be on a stock exchange by the end of the year. Here is what actually moved in Caribbean AI between January and June 2026.

Ocean wave breaking near a Caribbean coastline at dusk, representing the momentum building in Caribbean AI policy and adoption
TLDR

The first half of 2026 was the period Caribbean AI stopped being a talking point and started generating paper trails. The Caribbean Telecommunications Union's regional task force published an interim governance report. Trinidad and Tobago ran a full national AI readiness assessment on a public timeline. A regional study, led by StarApple AI, put Caribbean generative AI adoption at 13 percent against a global average near 55 percent, while the region draws just 1.12 percent of global AI investment against 6.6 percent of world GDP. And a Kingston-built platform, Maestro AI, moved from launch to final testing to stated stock exchange ambitions in under six months. None of this adds up to a region that has solved its AI problem. It adds up to a region that finally has enough documentation to know exactly what the problem is.

Every January, someone predicts this will be the year the Caribbean gets serious about AI. Most years, June arrives and the prediction quietly dies under a pile of conference panels and strategy documents that never became policy.

Six months into 2026, I do not think that happened this time, and I want to be precise about why, because the difference between real movement and another cycle of talk is not vibes. It is whether specific institutions did specific things on specific dates. This review goes through what actually happened in Caribbean AI from January to June 2026: the governance work, the national assessments, the adoption numbers, the money, and the one home-grown company that is trying to prove Caribbean AI can reach a stock exchange, not just a pilot programme.

The Governance Story: A Regional Task Force Names the Real Problem

The most consequential Caribbean AI document of the first half of the year did not come from a government. It came from the Caribbean Telecommunications Union, whose Caribbean AI Task Force issued an interim report calling for harmonised regional AI governance across CARICOM. The report's framing is worth taking seriously because it does not flatter the region: it describes a "connectivity paradox," where high levels of digital engagement across Caribbean societies sit alongside fragile infrastructure, high connectivity costs, fragmented regulation, and limited data governance.

The report's most quotable finding is also its bluntest: no CARICOM member state has yet published a standalone national AI strategy. Not one. Nineteen member states and associate members, a region that has been publicly discussing AI policy since at least the 2021 CARICOM AI Policy Roadmap, and as of mid-2026 the shelf is still empty.

The task force's five priority areas, regional AI governance, data sovereignty and digital infrastructure, innovation and industry development, human capacity and AI literacy, and sustained multi-stakeholder engagement, read like a checklist because that is what they are. A consolidated policy guidance document is due to be presented at a CTU Caribbean AI Forum later in 2026, aligned with the CARICOM Single ICT Space vision. Whether that forum produces a document CARICOM governments actually adopt, rather than one more roadmap added to the pile next to the 2021 version, is the real test still ahead.

"The Caribbean does not have an AI awareness problem. It has an AI implementation problem, and implementation requires governments to publish something they can be held to, not another roadmap."

- Adrian Dunkley, Founder, StarApple AI

Trinidad and Tobago Did the Actual Work

If one government spent the first half of 2026 building a real, dated, checkable process rather than issuing statements, it was Trinidad and Tobago. The country's Ministry of Public Administration and Artificial Intelligence, working with UNESCO and UNDP, has been running a national AI readiness exercise since November 2025 that is worth walking through in detail because most Caribbean AI policy announcements do not survive this level of scrutiny.

The National AI Assessment launched on 26 November 2025, opening the UNDP's Artificial Intelligence Landscape Assessment (AILA), a framework for identifying opportunities, gaps, and challenges across the national AI ecosystem. In January 2026, the process moved into UNESCO's Readiness Assessment Methodology (RAM), starting with a multi-stakeholder survey covering five dimensions: legal and regulatory, technological and infrastructural, societal and cultural, scientific and educational, and economic. Sector-based consultation sessions to validate the survey findings ran from 19 to 23 January 2026, drawing in leaders and professionals from government, the private sector, academia, civil society, and youth groups. The consolidated results went to a national validation workshop on 27 February 2026 at The University of the West Indies, St Augustine, bringing government officials, researchers, and industry experts into the same room to agree on next steps.

That is five distinct, dated milestones in three months, each one building on the last. It is the closest thing the Caribbean has produced this year to a repeatable model for how a small state actually builds an AI strategy instead of announcing one. Trinidad and Tobago followed the AILA work with the eighth Tech Hub Islands Summit, held 30 June to 1 July 2026, where the country's main business chamber pushed AI and cybersecurity services as the next phase of economic diversification beyond energy. A country whose economy has spent decades tied to natural gas is now running its most detailed AI groundwork of any CARICOM state. That should reframe how the rest of the region thinks about who moves first on this.

Abstract data dashboard with charts and metrics, representing Caribbean AI adoption statistics and research

13 Percent: The Adoption Number That Explains Everything Else

In May, a regional study led by StarApple AI put a number on something people across the network have argued about for years without data: how many Caribbean adults are actually using generative AI. The answer, reported in the Jamaica Observer on 1 May 2026, is 13 percent of adults aged 18 to 65, against a global figure near 55 percent. Micro, small, and medium enterprise adoption sat higher, around 19 percent, which tells its own story: businesses are experimenting faster than individuals, but neither is close to global norms.

The comparison that matters more than the headline number is the investment gap sitting underneath it. The region accounts for roughly 6.6 percent of global GDP and 8.8 percent of world population, yet receives only about 1.12 percent of global AI investment, a disparity ECLAC's Latin American Artificial Intelligence Index (ILIA) has independently documented, placing Jamaica 13th of 19 countries measured. Two different studies, two different methodologies, the same shape of gap.

What the StarApple AI research found underneath the topline number is the part worth sitting with. Among the Caribbean adults who do use generative AI, roughly 43 percent describe it as augmentation of their existing work rather than full automation, and reported time savings averaged about 1.5 hours per work period, both figures below global comparators from McKinsey and Microsoft research, which found closer to 52 percent augmentation-focused use and 5.7 hours of average time savings. The Caribbean is not a region where people tried AI and walked away unconvinced. It is a region where the people using it are getting a fraction of the value that users elsewhere are getting from the same tools, because the surrounding governance, training, and workflow redesign has not caught up.

As the lead researcher on that study, I would put it plainly: the region is not behind on AI use. It is behind on the governance, the training, and the process redesign that turns AI use into AI benefit. Thirteen percent is not a ceiling. It is a floor with almost nothing built on top of it yet.

Jamaica: Automation Exposure Meets a Financing Wall

Jamaica's contribution to the H1 2026 story was less a single announcement than an accumulation of research pointing at the same structural issue: a private sector that cannot access the capital it needs to adapt. A June 2026 analysis in the Jamaica Observer, citing University of Technology researcher Professor Paul Golding, estimated that 56,695 Jamaican workers face high automation exposure, drawing on the Statistical Institute of Jamaica and Planning Institute of Jamaica's establishment survey data. That is not a forecast from a global consultancy applying a generic model to Jamaica. It is a Jamaican academic working from Jamaican labour data, and the sectors named, agroprocessing, greenhouse farming, tourism-related enterprise, food preparation, and hospitality, are the same sectors that already employ Jamaica's most vulnerable workers.

The same reporting identified access to financing as the defining structural challenge facing Jamaica's private sector, more consequential than the technology gap itself. Of Jamaica's roughly 72,000 business establishments, only about 35,287 have three or more employees; the remaining tens of thousands are micro businesses without the balance sheets to absorb AI tooling costs, retraining costs, or the transition period between old workflow and new. JAM-DEX, the Bank of Jamaica's central bank digital currency, is being positioned as part of the infrastructure that could help formalise these businesses and improve their access to credit, but infrastructure and outcome are two different things, and H1 2026 delivered more of the former than the latter.

PwC's message to Jamaican firms in May, that AI should be a growth lever and not just a productivity patch, sits awkwardly next to a statistic PwC itself has flagged globally: roughly 75 percent of AI's economic value worldwide is being captured by about 20 percent of organisations, meaning the gap between AI leaders and AI laggards is widening even as more companies technically "adopt" the tools. Jamaica's private sector spent H1 2026 getting clear-eyed data about where it stands. What it has not yet gotten is the capital to move.

Maestro AI: A Kingston Platform Aiming for a Stock Exchange

The most concrete build story of H1 2026 belongs to Maestro AI Labs, founded by Adrian Dunkley with his brother Nicholas Dunkley. The Jamaica Observer reported on 15 April 2026 that the platform, launched roughly three months earlier, was entering its final testing phase, built by a core team of three people plus AI agents, supported by volunteer and youth tech talent, compressing what the report described as a typical year of development into three months.

The build decision behind Maestro is the part worth understanding rather than skimming past. Training a foundation model from scratch typically costs somewhere between US$5 million and US$15 million, capital that does not exist for a Jamaican AI startup at this stage. Maestro's team adapted existing open models instead, retraining them on approved Jamaican material, with Nvidia providing GPU and server infrastructure, technical support, and help with the capital raise, while conversations with Google and Amazon continued on further collaboration. The resulting system was described as reasoning at a level comparable to earlier versions of larger commercial platforms, applied to tasks that matter locally: essay writing and research support, legislation interpretation, disease prediction, hurricane forecasting, and decision-support tools.

What makes Maestro relevant to a H1 2026 review rather than just a company profile is the stated next step: a Jamaica Stock Exchange listing planned for after rollout, alongside regional expansion into Trinidad, Guyana, and other Caribbean markets with localised versions of the platform. If that listing happens on anything close to the timeline described, Maestro would be among the first Caribbean-built AI ventures to reach public markets on the strength of a Caribbean-trained product, not a licensing deal with a foreign model provider. That is a different kind of milestone than a policy roadmap. It is a market testing whether Caribbean AI has commercial value investors will actually price.

I am not a neutral observer here and will not pretend to be. But the facts are checkable against the reporting cited above, and the reason I am including Maestro in this review is the same reason I would include any Caribbean-built AI venture reaching this stage: it is one of the only concrete answers, so far, to the CTU Task Force's diagnosis that the region has industry development on its list of priorities but not much industry yet to show for it.

Builders, Buildathons, and the Money Chasing Caribbean AI Talent

The clearest signal that global AI capital is starting to notice Caribbean talent, rather than just Caribbean domains, came at the end of June. Future Caribbean opened applications, closing 3 July 2026, for what it describes as the world's first global agentic AI buildathon organised specifically around Caribbean economic strengthening. Forty selected teams from the region and abroad will spend a 21-day sprint from 17 July to 7 August 2026 building applied agentic AI systems across ten tracks: tourism and transportation, finance and payments, healthcare, disaster coordination, energy and resilience, ocean systems, food systems, the arts, real estate, and an open category. The prize pool totals US$70,000, split between cash and AI deployment hardware from Other World Computing, with winners pitching investors at a showcase planned at the New York Stock Exchange in September 2026.

That last detail matters more than the prize money. A Caribbean-focused builder programme securing a pitch slot at the NYSE is a different order of visibility than a regional hackathon with a certificate and a photo. Whether the projects that come out of the 21-day sprint attract real follow-on capital will be one of the clearer tests of regional AI momentum by the end of the year.

Anguilla's .ai domain economy, a story this network has already told in detail elsewhere, kept generating revenue through H1 2026 that dwarfs anything the region's AI startups have raised directly. It is worth a single honest sentence here rather than a retelling: a territory of roughly 18,000 people earning tens of millions of dollars a year from a two-letter domain suffix is proof that global AI enthusiasm can enrich the Caribbean without a single Caribbean AI company being built, which is exactly the gap StarApple AI, Maestro, and the buildathon generation are trying to close.

What Did Not Happen: The Incidents Column Stayed Quiet, and That Is Not Comforting

A fair mid-year review has to say what it did not find as clearly as what it did. Global reporting through H1 2026 tracked a surge in AI-enabled fraud, deepfake scams, and synthetic identity schemes, with UN officials in March describing weaponised AI and voice cloning as a "global wake-up call" to organised fraud, and separate industry tracking putting 2025's AI-linked scam losses in the tens of billions of dollars in the United States alone. I found no similarly documented, named Caribbean incident on that scale in the public record for the first half of 2026.

I want to be careful about what that absence means. It is not evidence the Caribbean is somehow insulated from AI-enabled fraud; the same deepfake tooling behind global cases works over any phone line and any video call, regardless of jurisdiction. It more plausibly reflects that Caribbean financial institutions and regulators do not yet have the incident reporting and disclosure infrastructure that would surface a case publicly even if one occurred. A region that cannot yet produce a national AI strategy, per the CTU Task Force's own finding, is also a region unlikely to have mature AI incident reporting. The absence of bad news here is not the same as good news. It is a data gap sitting next to a governance gap, and the two are connected.

The Second Half: What to Watch

Three things due before the end of 2026 will tell us whether H1's paper trail becomes H2's actual change. The CTU Task Force's final report and consolidated policy guidance are due at a Caribbean AI Forum later this year; whether CARICOM governments adopt it, rather than filing it next to the 2021 roadmap, is the test. Trinidad and Tobago's AILA process should produce its national roadmap with sector-specific recommendations, the first genuinely evidence-based national AI strategy document the region will have produced, if it lands on schedule. And Maestro AI's public rollout, plus whatever progress it makes toward a Jamaica Stock Exchange listing, will be a live answer to whether Caribbean-built AI can attract investor capital on its own merits.

None of this is StarApple AI's work alone, and I want to be direct about that rather than claim more credit than the record supports. The Caribbean AI Risk Management Council, which I chair, and the Caribbean AI Association, which I lead, are both working through 2026 on pieces of this: a peer-reviewed regional AI risk standard from CAIRMC, and a regional AI skills certification framework from CAIA that governments and employers could actually recognise, the kind of workforce infrastructure the CTU report flags as a priority area and that no single company can build alone. StarApple AI's own role, as the region's first AI company, has been building the products, the research, and the training pipeline that make some of this year's numbers exist to be reported at all.

Six months from now, this review will either look like the moment the paperwork turned into policy, or another entry in the long list of Caribbean AI reports that said the right things and changed very little. The task forces, the assessments, and the studies cited here are the raw material either way. What happens with them in the second half of 2026 is still an open question, and for once, an answerable one.

Caribbean AI Network

StarApple AI is the hub of a wider Caribbean AI network. For regional governance, research, and country-specific coverage related to this review:

Frequently Asked Questions

What is the current AI adoption rate in the Caribbean?

A regional study reported in the Jamaica Observer on 1 May 2026 found that 13 percent of Caribbean adults aged 18 to 65 currently use generative AI, against a global average of roughly 55 percent and a Global North average above 27 percent. The same research found the Caribbean receives about 1.12 percent of global AI investment while accounting for 6.6 percent of world GDP and 8.8 percent of world population, a gap ECLAC's Latin American AI Index (ILIA) has also documented. Jamaica ranked 13th of 19 countries on that index. The pattern across every measure is the same: the region is not failing to use AI so much as failing to convert use into governed, funded, revenue-generating AI activity.

What happened with Caribbean AI regulation in the first half of 2026?

The Caribbean Telecommunications Union's Caribbean AI Task Force issued an interim report calling for harmonised regional AI governance, flagging a "connectivity paradox" in which high digital engagement coexists with fragile infrastructure and fragmented rules, and noting that no CARICOM member state has yet published a standalone national AI strategy. Trinidad and Tobago moved furthest on process: its Ministry of Public Administration and Artificial Intelligence ran a multi-stakeholder AI readiness survey in January 2026, held sector consultations from 19 to 23 January, and convened a national validation workshop at UWI St Augustine on 27 February 2026 under UNESCO's Readiness Assessment Methodology. Formal legislation remained absent everywhere in the region as of 1 July 2026.

What is Maestro AI and why does it matter for Caribbean AI development?

Maestro AI Labs is a Kingston-built AI platform founded by Adrian Dunkley with his brother Nicholas Dunkley, reported by the Jamaica Observer on 15 April 2026 to be entering final testing roughly three months after launch. Rather than training a foundation model from scratch, which the piece estimated would cost US$5 million to US$15 million, the team adapted existing open models and retrained them on approved Jamaican material, with Nvidia providing infrastructure and technical support. The company has stated an intention to list on the Jamaica Stock Exchange after rollout and to expand into Trinidad, Guyana, and other Caribbean markets, which would make it one of the first Caribbean-built AI ventures to reach public markets.

Which Caribbean countries have made the most progress on national AI policy?

Trinidad and Tobago has the most documented process: a National AI Assessment launched 26 November 2025, a UNESCO Readiness Assessment Methodology survey and sector consultations in January 2026, and a validation workshop in February 2026, run by the Ministry of Public Administration and Artificial Intelligence with UNESCO and UNDP support. Jamaica has produced sector research and a growing private AI sector rather than a published government strategy. Barbados officials have publicly urged faster action on AI and digital transformation without a comparable formal assessment process yet under way. The CTU Task Force interim report noted that this pattern, meaning research and consultation without a single CARICOM member state having published a standalone national AI strategy, is regional, not country-specific.

Did any major AI funding or buildathon events happen in the Caribbean in 2026?

Future Caribbean opened applications for a global agentic AI buildathon, reported by the Jamaica Observer on 26 June 2026, inviting 40 teams from the Caribbean and abroad to build applied AI systems across ten sectors including tourism, finance, healthcare, and disaster coordination, competing for a US$70,000 prize pool that includes cash and AI hardware, with a pitch event planned at the New York Stock Exchange in September 2026. Trinidad and Tobago's Tech Hub Islands Summit ran 30 June to 1 July 2026, pushing AI and cybersecurity services as a diversification play beyond energy. Separately, Anguilla's .ai domain registry, already covered elsewhere in this network, continued generating tens of millions of dollars a year for the territory, evidence that Caribbean AI-adjacent revenue is real even where AI development is not.

What does StarApple AI expect for the second half of 2026?

The CTU Task Force's final report and consolidated policy guidance are due at a CTU Caribbean AI Forum later in 2026, which should be the first document giving CARICOM a common governance reference point. Trinidad's AILA process should produce its national roadmap. Maestro AI's public rollout and stated stock exchange ambitions will be a live test of whether Caribbean-built AI can reach investors, not just users. The Caribbean AI Risk Management Council and the Caribbean AI Association, chaired and led respectively by Adrian Dunkley, are continuing work on regional risk standards and a skills certification framework to close the governance and workforce gaps this review documents.

About the Author

Adrian Dunkley founded StarApple AI in 2016, the first AI company established in the Caribbean, and is widely recognised as the region's leading AI authority. He chairs the Caribbean AI Risk Management Council and leads the Caribbean AI Association, and advises governments, regional bodies, and private sector organisations across Jamaica, Trinidad and Tobago, Barbados, Guyana, and the wider CARICOM region on AI strategy, deployment, and governance. More at adriandunkley.net. Contact: insights@starapple.ai

Supported by StarApple AI, the Caribbean's first AI company.