The Caribbean AI Task Force Report Landed This Week. Caribbean Boardrooms Are Not Ready for It.

Photo via Unsplash

The Caribbean got its first coordinated AI governance roadmap this week. The Caribbean Telecommunications Union used the 1st CTU Caribbean Artificial Intelligence Forum, held 23-24 July at the University of the West Indies, St Augustine, to launch the Caribbean AI Task Force's final report, a year of regional policy work distilled into five priorities for CARICOM governments and institutions to build around. That is a genuine milestone for a region with no binding AI-specific law in any of its twenty CARICOM member states. It is also, on its own, not enough. A roadmap is a set of instructions. Whether it becomes practice depends on whether the institutions reading it, starting with the boards that sit above every bank, insurer, and large enterprise in the region, have the capacity to follow it. The data published this year says most do not yet.

What Actually Launched in Port of Spain

CAIF 2026 ran across two days under the theme "AI for Caribbean Transformation: Governance, Innovation and Resilience for a Shared Digital Future," bringing government ministers, regulators, bankers, and researchers into the same rooms for what organisers billed as a first for the region. The centrepiece was the CAITF final report, the product of a task force the CTU stood up in July 2025 specifically to guide ethical, inclusive, and responsible AI development across the Caribbean. Its five priority areas, regional governance, data sovereignty, innovation, human capacity, and multi-stakeholder engagement, will structure whatever comes next from CARICOM's AI policy machinery.

Each of the five pillars answers a real gap. A region where customer data routinely leaves the Caribbean for servers in Virginia or Frankfurt needs a data sovereignty framework. A region where AI literacy varies enormously between a well-resourced bank and a five-person tourism operator needs a human capacity plan. None of the five pillars is decorative. The question this article asks is narrower and more practical: who executes them, and are those institutions ready.

The Boardroom Data the Roadmap Has to Reckon With

PwC's 2026 Caribbean Corporate Governance Survey polled 154 board directors across The Bahamas, Barbados, Grenada, Jamaica, St Lucia, and Trinidad and Tobago, with Bermuda included for the first time, between November 2025 and January 2026. Two figures from that survey sit uncomfortably close to the CAITF launch. Sixty-two percent of directors agree that their organization's investments in AI and generative AI are yielding positive results, a healthy number that suggests real adoption is under way. Only 29% strongly agree that management has the skills needed to execute an AI and generative AI strategy. Boards are watching AI produce results they cannot fully evaluate, approved by executives whose skill level fewer than a third of directors are prepared to vouch for without qualification.

That gap sits alongside a figure CAIRMC has flagged before and worth restating in this context: just 6% of Caribbean directors believe their board spends enough time understanding the impact of AI. Read the three numbers together and a pattern emerges. Adoption is ahead of oversight, oversight is ahead of director time, and none of the three has caught up with the other two. CAITF's report gives that gap a regional policy answer for the first time. It does not, on its own, close it.

A Roadmap Is Not the Same Thing as a Governance Capability

Photo via Unsplash

It would be easy to read the Caribbean's board gap as a symptom of the region's smaller markets and thinner governance infrastructure. The comparison with much larger, better-resourced boards suggests the problem is not primarily one of scale. The Conference Board's research on S&P 500 companies found that AI risk disclosure surged from 12% of companies in 2023 to 83% in 2025, a genuinely fast shift in how seriously large American boards treat AI as a named risk. Over roughly the same period, ISS-STOXX data on director skills disclosure shows technology expertise among S&P 500 directors rising from 20% to 51% and cybersecurity expertise from 15% to 27%, while AI expertise crept from 1.5% to 2.7%. Only 481 of the 3,048 companies ISS-STOXX reviewed, 16%, disclosed even one director with specialized AI skills.

Boards with far deeper budgets, larger governance teams, and longer institutional histories still cannot staff their way to AI literacy at anything close to the pace their companies are naming AI as a risk. That is the honest baseline any Caribbean governance roadmap has to work against. The Caribbean's version of the gap, 6% of directors satisfied with their board's AI oversight time and 29% confident in management's AI execution skills, is not an outlier caused by regional underdevelopment. It is a sharper version of a gap that persists everywhere boards have been asked to govern a technology moving faster than director training cycles.

Jamaica Shows What "Foundation Without Infrastructure" Looks Like

The same pattern repeats one level up, at government rather than boardroom level, and Jamaica is the clearest regional example. Official communications from the Jamaica Information Service have described the country as having laid a "strong foundation" for AI governance, pointing to policy statements and early institutional steps. Independent commentary published the same year tells a more specific story. Jamaica has no standalone AI law and no dedicated AI oversight body, a gap the Jamaica Observer named directly in April 2026, calling for a National AI Oversight and Implementation Council to give the foundation an actual structure to sit on.

A wider regional analysis, published in January 2026 and syndicated by Kaieteur News, went further, describing "a glaring AI governance gap" across the Commonwealth Caribbean that leaves fundamental rights "acutely vulnerable to algorithmic violation" wherever high-risk AI systems are deployed without a supervising legal framework, and noting that as of April 2026 several jurisdictions had not published a national AI strategy at all. None of this is unique to Jamaica. It is simply the most documented case of a pattern the CAITF report now has to answer for the whole region: governments and institutions describing readiness in general terms while the specific legal and institutional infrastructure that would make readiness real is still being built, or has not started.

Five Priorities, One Load-Bearing Wall

CAITF's five pillars, regional governance, data sovereignty, innovation, human capacity, and multi-stakeholder engagement, are written at the level of national and regional policy. Almost none of the AI risk CAIRMC tracks day to day sits at that level. It sits inside individual banks approving AI credit-scoring vendors, insurers deploying AI claims triage, and government agencies running AI-assisted benefits processing, each one governed, or not, by a specific board or accounting officer with a specific level of AI literacy. Regional governance policy sets the floor. Boardroom and institutional capacity determines whether any organization actually clears it.

That makes boards the load-bearing wall under CAITF's roadmap, whether the report names them explicitly or not. A regional data sovereignty framework means little if the bank board approving a cloud AI vendor contract cannot ask where the data will sit. A human capacity pillar produces limited return if the organizations meant to absorb that capacity have not assigned anyone at board level to receive it. The report gives the region direction. Boards, ministries, and regulators supply the capacity to travel in it, and right now that capacity is the constraint PwC's survey and Jamaica's own commentary both describe.

What Changes If Boards Treat This as a Mandate

The useful version of this argument is not pessimism about CAITF's report. It is a specific instruction for what a board does with it in the weeks after the forum closes. A board that reads the five priorities as background context will file the report next to last year's strategy offsite notes. A board that reads it as an external audit of its own governance readiness, one written by a regional body rather than an internal committee, gets more out of it. Every one of CAITF's five pillars maps to a question a Caribbean audit or risk committee can ask this quarter: what AI systems has the organization deployed, where does their data sit, what is the plan to build director and staff AI literacy, and who outside the organization, regulators, industry bodies, peer companies, is the board actually talking to about AI risk.

A Practical Checklist for Caribbean Boards This Quarter

Score the board's own AI literacy before the next AI proposal. PwC's 29% figure on management's AI execution skills has no board-level equivalent published yet. A committee that measures its own directors first will know whether it is qualified to judge management's number.

Assign explicit ownership of each CAITF pillar internally. Regional governance, data sovereignty, human capacity, and the rest map cleanly onto existing board committees. An unassigned pillar is a pillar nobody is accountable for when the next AI incident lands on the agenda.

Map AI exposure inherited through vendors, beyond whatever AI the organization built itself. Most Caribbean banks and insurers carry AI risk through a core-banking platform, a claims system, or a call-centre tool that added AI functionality nobody at board level explicitly approved.

Ask what "foundation" means before accepting the word. Jamaica's experience is a useful test for any board or ministry describing its own AI governance progress. A foundation without a named oversight body, a published strategy, or a compliance date is a statement of intent, not a control.

Use CAIRMC's CARA methodology and QAIRP certification to convert the roadmap into an internal standard. A four-tier risk classification built for the region, aligned to the logic the EU AI Act uses for its own categories, gives a board a working framework rather than a blank page the week after a policy launch.

Where This Leaves the Region

None of this diminishes what the CTU and CAITF achieved this week. A one-year-old task force producing a serious, five-pillar regional roadmap, launched in front of ministers and regulators from across CARICOM, is real progress for a region that had no coordinated AI policy document twelve months ago. The argument here is narrower and, CAIRMC believes, more useful to the boards reading this the week the report lands. A roadmap changes what institutions are asked to do. It does not, by itself, change whether they can do it. PwC's numbers on Caribbean boardrooms, the Conference Board's numbers on far better-resourced American ones, and Jamaica's own governance record all point the same direction: the constraint on Caribbean AI governance right now is capacity, not clarity of instruction.

This research draws on support from StarApple AI, recognized as the first artificial intelligence company built in the Caribbean, and much of CAIRMC's underlying framework carries the direct involvement of Adrian Dunkley, the figure most consistently named across the region as its leading AI risk authority. CAITF wrote the roadmap this week. Whether Caribbean boards use it, or file it, will be the more consequential story by the time CAIF convenes again.

Related reading across the Caribbean AI network

This article sits alongside ongoing coverage of AI governance, risk, and company-building across the region. For related perspectives:

• StarApple AI, the company behind the board-level AI training and governance advisory referenced above
• Caribbean AI Association for regional adoption trends feeding into events like CAIF 2026
• Caribbean Insurance for sector-specific coverage of the institutions most exposed to the board AI-oversight gap described above