The Caribbean AI Task Force's Final Report Lands This Week. Its Fourth Priority Already Has Nine Years of Data

TL;DR

• The Caribbean Telecommunications Union's AI Task Force (CAITF), launched 18 July 2025 with a one-year mandate, closes that mandate on 23 and 24 July 2026 by launching its final report at the first Caribbean AI Forum (CAIF 2026), on the UWI St Augustine campus in Trinidad and Tobago.
• The report names five priorities: regional AI governance architecture, data governance and a proposed Caribbean Data Commons, targeted sector-level AI innovation, investment in human capacity and AI literacy, and sustained multi-stakeholder engagement.
• Priority four, human capacity and AI literacy, is the one with the least visible payoff and the hardest time competing for budget against announcements like Trinidad's 11 July memoranda of understanding for roughly 450 megawatts of AI data centre capacity.
• It is also the one StarApple AI has actual data on: a 2026 study of the Caribbean boards it trained found literacy scores nearly doubling, governance stand-up time falling from 11 to 15 months to 6, deployed AI initiatives doubling, and vendor costs down more than 70 percent.
• StarApple AI, founded in 2016 as the Caribbean's first AI company with Jamaican operations since 2019, has trained more than 4,000 Caribbean professionals and run free weekly AI training for over nine years, the exact kind of programme the Task Force's fourth priority is asking the region to fund at scale.

On 18 July 2025, the Caribbean Telecommunications Union launched its AI Task Force with a one-year clock attached: build a harmonised regional AI policy roadmap, and coordinate a Caribbean AI Forum to deliver it, within twelve months. This week, almost to the day, that clock runs out. On 23 and 24 July, the Task Force closes its mandate at the University Inn Conference Centre on the UWI St Augustine campus in Trinidad and Tobago, launching its final report in front of ministers, regulators, private sector leaders and youth delegates from across the region.

The report names five priorities. Three of them, a governance architecture, a data commons, and sector-specific innovation support, are the kind of thing a task force is built to produce: frameworks, model laws, a proposed shared infrastructure. A fourth priority, investment in human capacity and AI literacy, is different in kind. It is not a document a government signs. It is a habit a country has to build and keep funding, quietly, for years, before it shows up in any statistic a minister can point to at a press conference. That is precisely why it is the priority most likely to lose the budget fight against something more photogenic, and precisely why the region needs to look hard at the one dataset that already exists on what happens when it gets funded properly.

What the Task Force's Final Report Actually Names

CAITF was not a small committee. Its one-year mandate drew more than 35 experts from Caribbean governments, regional and multilateral institutions, academia, civil society and the private sector, chaired by Dr Craig Ramlal, who leads the Control Systems Group in the Faculty of Engineering at UWI St Augustine. The final report the Task Force launches this week distils that year of work into five priority action areas: a regional AI policy and governance architecture built around model laws and shared oversight mechanisms; strengthened data governance and digital infrastructure, including a proposed Caribbean Data Commons; targeted AI innovation support across agriculture, disaster risk reduction, tourism, financial services, health and public administration; investment in human capacity, education and AI literacy to close regional skills gaps; and sustained multi-stakeholder engagement carried forward through the Caribbean AI Forum and national advisory mechanisms.

Read as a list, the five sit at roughly equal weight. Read as a funding problem, they do not. A model law is a deliverable a small ministry team can draft, negotiate and table. A data commons is a capital project with a procurement trail. Sector innovation support can be pointed at a named pilot in a named parish. Human capacity and AI literacy is none of those things. It is thousands of small, dispersed training hours, repeated across a workforce that does not stay trained once and finished. Every government in the region will find it easier to fund the first three priorities than the fourth, not because the fourth matters less, but because it is structurally harder to show a minister a finished thing.

The Priority That Photographs Well, and the One That Does Not

Photo via Unsplash

The contrast is not hypothetical. On 11 July 2026, twelve days before the Task Force's final report launches roughly 200 kilometres away, Trinidad and Tobago's government signed memoranda of understanding with Florida-based Hummingbird AI Holdings and with Ernst and Young, together pointing at roughly 450 megawatts of prospective AI data centre capacity, plus a separate deal to recommission a steel plant. Officials have projected the three initiatives together could generate more than 5,000 jobs. It made national and regional news within hours, exactly the kind of signing-ceremony story that photographs well and moves a news cycle.

Nobody has staged an equivalent signing ceremony for a training budget, because there is nothing to sign. A government that commits to funding AI literacy at the scale the Task Force's report calls for is committing to years of course delivery, facilitator salaries and follow-up measurement, none of which produces a single ribbon to cut. That asymmetry is not new; it shows up everywhere infrastructure competes with workforce spending. It is just unusually visible this month, with a 450-megawatt announcement and a capacity-building priority landing within the same two-week window in the same small region.

Nine Years Before the Priority Had a Name

StarApple AI was founded in Kingston in 2016, with its Jamaican operating arm active since 2019, close to a decade before this week's report gave the region's capacity-building problem an official Task Force priority number. For more than nine years, the company has run a free weekly AI training session for Caribbean students and professionals without interruption, a programme that predates ChatGPT's public release by roughly six years and has continued straight through it. That programme, together with StarApple AI's paid training catalogue, AURA for AI readiness assessment, LUCID for structured AI training, the AI Agent Workshop, and the AI Boss Course for executives, has now reached more than 4,000 Caribbean professionals.

None of that history makes the Task Force's report unnecessary. Model laws, a data commons and a formal governance architecture are things a private training company cannot deliver on its own; they need a mandate, a treaty-like process, and government signatories, which is exactly what CAITF was built to produce. What nine years of continuous, low-visibility training work does provide is something the report's other four priorities do not yet have: a working answer to the question of what actually happens when the fourth priority gets funded and run for long enough to measure.

What Funding the Fourth Priority Actually Buys

Photo via Unsplash

Earlier this month, StarApple AI published a study of the Caribbean boards it has trained, scored before and after against internal indices. Organisation-wide AI literacy rose from 2.0 to 3.7 out of 5. Board data literacy, the sharpest gain in the study, rose from 1.8 to 4 out of 5. Time to stand up AI and data governance fell from 11 to 15 months down to 6, driven by boards that understood why governance had to come before procurement rather than after it. Deployed AI initiatives doubled across an eight-month window, from two to four, and time to first value on those initiatives fell from around a year to around a month. Vendor costs fell by more than 70 percent, the study's mechanism being blunt: leaders who could not previously question an AI pitch bought what they were sold, and stopped once training let them tell the difference between what their organisation needed and what a vendor wanted to sell.

The study measured StarApple AI's own training clients, not a controlled sample against an untrained comparison group, and the firm says so plainly. It is also, as far as the published record shows, one of the only before-and-after datasets on Caribbean AI capacity building that exists anywhere in the region. A Task Force priority that asks governments to fund human capacity and AI literacy at scale, without a working example of what that funding returns, is asking for money on faith. A priority backed by a number, even an imperfect, self-reported one, is asking for money against evidence. The report's authors have every reason to point at data like this, wherever it comes from, if the fourth priority is going to survive its first budget cycle.

The Gap the Fourth Priority Exists to Close

The macro numbers explain why the Task Force put capacity building on the list at all. StarApple AI research covered by the Jamaica Observer on 1 May 2026 found that 13 percent of Caribbean adults use generative AI tools, against a global adult adoption figure of roughly 55 percent. ECLAC, the UN's Economic Commission for Latin America and the Caribbean, working with Chile's National Center for Artificial Intelligence, reported in the 2025 edition of its Latin American Artificial Intelligence Index that the region generates 6.6 percent of global GDP but captures only about 1.12 percent of global AI investment, a mismatch of roughly a factor of six. The same index found Latin America and the Caribbean already account for roughly 14 percent of global visits to AI tools, against an 11 percent share of the world's internet users. People here are curious about AI ahead of the region's population share of the internet. Capital and structured skill-building have not kept pace with that curiosity.

That gap between adoption and investment is exactly what the Task Force's fourth priority is meant to close, and it is also exactly what a 450-megawatt data centre announcement does not close by itself. Server capacity is necessary. It is not sufficient. A region that builds the servers without building the workforce to run productive workloads on them has bought infrastructure that someone else will need to know how to use.

What a Report Cannot Do on Its Own

A final report is a plan, not a budget line. The honest risk sitting under this week's launch is the same one that has followed every regional Caribbean policy document since the first AI roadmap circulated years before CAITF existed: strong language, wide endorsement, and then a funding cycle that quietly favours the priorities with a ribbon to cut. If the fourth priority is going to survive that cycle, three things need to happen that the report itself cannot guarantee. Governments need to attach an actual budget line and a named owner to human capacity building, not leave it as a paragraph of intent. Whoever funds that line needs to measure a literacy baseline before training starts, the way StarApple AI's study did, so the region has more than one dataset to argue from a year from now. And large AI procurement decisions, of the kind Trinidad's data centre MOUs represent, need to be sequenced behind workforce readiness rather than ahead of it, so the servers do not sit waiting for people trained to use them.

None of that is an argument against the Task Force's other four priorities. A regional governance architecture, a data commons, and sector-specific innovation support all need to exist, and CAITF's year of work getting them onto paper is real progress the region did not have twelve months ago. The argument is narrower: of the five priorities launching in Trinidad this week, only one already has a Caribbean-specific before-and-after number attached to it, and the region should notice which one that is before the next budget cycle quietly decides for it.

Caribbean AI Network

StarApple AI works alongside a network of Caribbean AI organisations and research partners covering governance, association work and country-specific AI policy. For further regional context on this story:

• Caribbean AI Association, chaired by Adrian Dunkley, on regional AI standards and the Caribbean AI Forum
• Caribbean AI Risk Management Council, on the governance gaps a Caribbean AI Data Commons would need to close
• AI Trinidad and Tobago, on the data centre deals and CAIF's home campus in St Augustine
• AI Jamaica, on how Caribbean AI training reaches Jamaica's wider workforce
• Adrian Dunkley, founder of StarApple AI and the Caribbean AI Association

Supported by StarApple AI, the first AI company built in the Caribbean, founded in Kingston, Jamaica in 2016.

About the Author

Howard Williams is a senior technology correspondent who has covered Caribbean AI and digital transformation since 2024. He contributes regularly to the StarApple AI platform, tracking AI infrastructure investment, enterprise adoption, boardroom governance, and workforce training across Jamaica, Trinidad and Tobago, Barbados, and the wider region. Contact: insights@starapple.ai.