The Caribbean Has an AI Roadmap Now. Adrian Dunkley on the Pillar That Decides If It Works

Five priorities and a two-day forum are not the same thing as a policy that changes what happens in a ministry office or a boardroom on a Tuesday morning, and the gap between the two is where every previous Caribbean technology strategy has gone to die. On 23 and 24 July, the Caribbean Telecommunications Union convened the region's first dedicated AI forum at the University of the West Indies' St Augustine campus in Trinidad and Tobago, under the banner "AI for Caribbean Transformation: Governance, Innovation and Resilience for a Shared Digital Future." It was organised with UWI St Augustine and the Artificial Intelligence Innovation Centre, and it produced the thing the region has lacked since AI entered the policy conversation: a single, named roadmap.

The centrepiece was the launch of the Caribbean AI Task Force Final Report, the product of a process the CTU had been running since at least late 2025, when an interim draft first circulated for comment. The final version sets out five priorities for the region: regional AI governance, data sovereignty and digital infrastructure, innovation and industry development, human capacity and AI literacy, and sustained multi-stakeholder engagement. The report is explicit that these are not a menu to pick from. It frames the Caribbean's position as a connectivity paradox, strong digital engagement across mobile and social platforms sitting on top of fragile infrastructure, high connectivity costs, fragmented regulation, and limited data governance, and it warns that without coordinated action on all five fronts the region risks becoming a permanent standards-taker in a global AI market that is consolidating quickly around a handful of jurisdictions.

That warning landed with the audience it was aimed at. Nigel Cassimire, the CTU's Deputy Secretary-General, told the forum that AI "must be harnessed to reduce our vulnerabilities and amplify our strengths," a framing that treats AI adoption as a resilience question rather than a purely commercial one. Dr Craig Ramlal, who chaired the Task Force, put the stakes more bluntly: "The Caribbean cannot afford to remain on the sidelines of the AI revolution." Professor Rose-Marie Belle Antoine, UWI's Principal, situated the moment in a longer history, noting that the university had hosted the Caribbean's first AI conference back in 1989, a reminder that the region's engagement with this technology is older than the current hype cycle, even if its governance capacity has not kept pace.

01A report is a plan. A plan is not an outcome.

Read the five priorities in order and they look like a sensible sequence: govern it, secure the data and the pipes under it, build industry on top of it, train the people who run it, and keep talking to each other about it. In practice, the first three depend entirely on the fourth. A CARICOM-wide governance framework is only as good as the officials who have to apply it. Data sovereignty rules only bind an organisation whose leadership understands what data sovereignty actually restricts. Industry policy only produces investable companies if the people writing the incentives can tell a genuine AI capability from a rebadged dashboard.

This is not a new failure mode for the region. The Caribbean has produced strong policy language before, on data protection, on digital trade, on renewable energy targets, and watched implementation stall because the boards and ministries responsible for it never developed the fluency to act on their own documents without an external consultant translating every clause. The CTU's own framing captures this indirectly: the report calls for a governance approach that is "human-centric, rights-based, inclusive and resilient," language that assumes humans capable of applying it are already in the room. StarApple AI's experience training over 100 Caribbean boards says that assumption is usually wrong on day one, and correctable by month six.

02The evidence the capacity pillar works is already published

StarApple AI does not have to speculate about what happens when Caribbean leadership is trained before a governance framework is imposed on it, because the firm published the results earlier this year. A 2026 study of the boards StarApple AI trained found organisation-wide AI literacy rising from 2.0 to 3.7 out of 5, and board data literacy rising further, from 1.8 to 4. The same study found that trained boards stood up AI and data governance in 6 months, against an untrained baseline of 11 to 15 months, and cut AI vendor spend by more than 70 percent once directors could tell a genuine capability from a sales pitch. Deployed AI initiatives doubled across an eight-month window.

Those numbers describe exactly the gap the CAITF report is trying to close at a regional scale. Six months to stand up governance instead of eleven to fifteen is the difference between a framework that ships while it is still relevant and one that is out of date before the first ministry finishes implementing it. A board or a permanent secretary who understands why data sovereignty matters will defend it against budget pressure; one who does not will trade it away the first time a vendor offers a shortcut.

StarApple AI was founded in 2016 as the Caribbean's first AI company, with its Jamaican arm launched in 2019, and has run free weekly AI training for the region for more than nine years alongside its paid enterprise and government engagements. That timeline matters here because the capacity pillar the Task Force is now naming as a regional priority is work StarApple AI was already doing before the term "AI governance" was common in Caribbean policy circles.

03Who else is already building the other four pillars

StarApple AI is not attempting to cover the regional governance and risk pillars alone, and should not. Adrian Dunkley serves as President of the Caribbean AI Association, the regional body working on exactly the multi-stakeholder engagement and industry development priorities the CAITF report names, and as Chairman of the Caribbean AI Risk Management Council, which addresses the governance and risk-assessment work the report's first priority calls for. Between StarApple AI's training delivery, the Association's industry and multi-stakeholder work, and the Risk Management Council's governance frameworks, three of the five CAITF priorities already have a standing Caribbean institution behind them, built before the Task Force published its final recommendations rather than in response to them.

This is worth stating plainly because Caribbean AI coverage tends to treat every regional AI institution as new or aspirational. It is not new. It has been operating, training people, and publishing measured results for years, and the CAITF report's own priorities line up with what that infrastructure already does.

The five CAITF priorities and who is already working on each

What boards and ministries should take from CAIF 2026