An Open Letter to the Heads of Government of CARICOM: Why AI Must Be a Priority Now

From the Caribbean AI Risk Management Council, addressed to the fifteen Heads of Government of the Caribbean Community, to the Conference of Heads of Government, the Bureau, and to the Caribbean public.

Honourable Heads of Government,

We write to you, in a spirit of regional partnership, on a matter that does not yet appear at the top of the CARICOM agenda but which, within the term of the current Conference, will reshape every sector that does. Artificial intelligence is already operating in our financial systems, our public services, our schools, our hospitals, our tourism platforms, and our information environments. It is doing so largely under the rules of jurisdictions that are not ours, on training data that is not ours, with vendor terms that are not ours, and at a pace that has outrun the policy machinery of even the largest member states. This letter is a respectful but urgent request that you collectively change that.

The Window That Is Closing

The European Union has passed the AI Act and is now implementing it. The United States is regulating AI through executive action and sector-specific rules. The People's Republic of China has issued binding regulations on generative AI and recommendation algorithms. The African Union has adopted a continental AI strategy. The OECD AI Principles are now the reference framework for forty-six member countries. The United Nations has established a high-level advisory body and a global AI governance dialogue.

The Caribbean Community has not yet adopted a regional AI strategy, a regional AI ethics framework, or a regional position on the international AI governance architecture being built without us. Several member states have begun individual national efforts. Jamaica has published consultative material. Barbados has signalled intent. Trinidad and Tobago has work in progress. The Cayman Islands and The Bahamas have engaged through their financial-sector regulators. These are useful starts. They are not, on their own, a CARICOM response.

The window in which we can shape the international AI governance regime, rather than inherit it, is closing. Major standards are being set this year and next. Caribbean voices are largely absent from those tables. The cost of absence is not symbolic. It is the regulatory shape that decides whether our financial services exports remain compliant, whether our students' data is exported lawfully, whether our citizens can challenge an AI decision made about them, and whether our regional cybersecurity exposure to AI-augmented attack is taken seriously by partner jurisdictions.

What We Are Asking For

We do not ask CARICOM to draft new technology. We ask CARICOM to make a small number of decisions that only Heads of Government can make.

First, we ask that AI be placed on the agenda of the next Conference of Heads of Government as a standing item, with a Lead Head of Government designated for the portfolio, in the same way that the Single Market, Climate Change, and Security have Lead Heads. The current absence of such a designation is the clearest single indication of the policy gap.

Second, we ask that the CARICOM Secretariat be directed to develop, within twelve months, a Regional AI Policy Framework. The framework should align with the UNESCO Recommendation on the Ethics of AI, which every CARICOM member has already adopted, and should articulate Caribbean-specific positions on training data sovereignty, automated-decision rights, AI in public services, AI in elections, AI in education, and AI in disaster management. The framework should not attempt to be a complete regulatory rulebook. It should be a coordinated regional position that member states can localise.

Third, we ask that a Regional AI Risk and Capability Assessment be commissioned. The Caribbean lacks reliable data on where AI is currently being used in our public and private sectors, by whom, on what data, and with what controls. Without that assessment, regional policy will be made in the dark. CAIRMC, the University of the West Indies, the University of Guyana, the Caribbean Telecommunications Union, and the Caribbean Centre of Excellence for Sustainable Livelihoods are well placed to support the assessment.

Fourth, we ask that CARICOM speak with one voice in the international AI governance forums. The UN AI Advisory Body, the OECD AI Policy Observatory, the AI Safety Summit process, and the regional financial-sector standards setters all have processes that are accepting input now. A coordinated Caribbean voice carries more weight than fifteen separate voices, particularly when many of our member states do not have the resident specialist capacity to staff these forums individually.

Fifth, we ask that CARICOM resource a Regional Capacity Programme on AI. This is not a request for large capital expenditure. It is a request for a sustained, modest annual commitment to train regional public servants, regulators, judges, and educators in AI literacy and AI governance. Without that capacity, every other recommendation in this letter will sit unimplemented.

Why a Regional Response, Not Fifteen National Responses

Several CARICOM member states have, understandably, asked why this work should be coordinated regionally rather than left to each Government. There are five answers.

The technical assets are shared. Most AI tools deployed in the Caribbean are provided by a small number of foreign vendors. Negotiating standards individually with those vendors leaves each member state with weaker terms than the region could secure collectively.

The talent base is shared. UWI graduates serve every member state. Regional regulators draw on a common pool. A regional AI capability strategy yields more return per dollar than fifteen parallel ones.

The risks are shared. A misclassification by a Caribbean customs AI agent in one jurisdiction creates litigation exposure in another. A deepfake campaign during an election in one member state propagates across the region's media in days. AI risk does not respect maritime boundaries.

The voice is shared. A Caribbean position at the UN, OECD, or IMO carries weight in proportion to the size and coherence of the bloc speaking. Fifteen national positions speak with less weight than a coordinated CARICOM position.

The history is shared. From the Caribbean Court of Justice to the Caribbean Public Health Agency to the Single Market, our region has built credible joint institutions where the cost of going alone was too high. AI governance now meets that threshold.

What the Region Stands to Lose

If CARICOM does not act, the following outcomes are not speculative. They are the trajectory we are already on.

Caribbean credit, insurance, hiring, and benefits decisions will increasingly be made by AI systems trained on data that does not reflect Caribbean populations. The bias will be statistical, the resulting harms will be specific, and the avenues for redress, in most member states, will be unclear.

Caribbean elections will face AI-augmented disinformation campaigns, deepfakes of candidates, and synthetic media at a scale that small electoral commissions cannot independently rebut. The OECS, in particular, with its frequent and high-turnout electoral cycles, is exposed.

Caribbean financial centres will face standards-driven exclusion from correspondent banking, financial messaging, and cross-border investment flows if regional AI governance does not meet the expectations of partner jurisdictions. The cost of AI under-regulation will mirror the cost of AML under-regulation a decade ago.

Caribbean educational outcomes will diverge from international peers if our schools and universities do not collectively adopt clear, enforceable AI-use frameworks. Brain sleep, as we have argued elsewhere, threatens to follow brain drain as a regional cognitive crisis.

Caribbean climate adaptation, increasingly dependent on AI-supported early warning systems, predictive modelling, and remote sensing, will be slower and less effective than it should be if regional AI capability does not catch up to the scientific opportunity. Hurricane Melissa, in 2026, should have been the wake-up.

What CAIRMC Will Do

The Caribbean AI Risk Management Council was constituted to support exactly this work. We commit, on the public record, to the following.

We will publish the Caribbean AI Risk Management Standard 2026 First Edition, freely available to all CARICOM member states, regulators, and public-sector bodies.

We will provide, without cost, capacity-building sessions for any CARICOM ministry, regulator, or judicial body that requests them, subject to scheduling.

We will represent Caribbean AI risk and governance interests in international forums, in coordination with regional institutions and at the invitation of member governments.

We will continue to publish open, accessible material aimed at Caribbean professionals, regulators, and citizens. AI literacy, written for our context, is the foundation on which any sound regional response will be built.

We do this in support of, not in substitution for, the work of Heads of Government, the CARICOM Secretariat, and the relevant regional institutions. We are conscious of our place. The decisions that this letter calls for are not ours to make.

The Caribbean Has Made Decisions Like This Before

Many of the regional institutions we now take for granted, the Single Market, the Caribbean Court of Justice, the Caribbean Disaster Emergency Management Agency, the Caribbean Public Health Agency, were the result of a small number of Heads of Government deciding, at a moment of regional pressure, that the cost of fragmented national action exceeded the political cost of regional coordination. AI is now at that moment.

We respectfully submit that the alternative to CARICOM acting on AI in 2026 is not the absence of AI in the Caribbean. It is the presence of AI in the Caribbean, on terms set elsewhere, that we will spend the next two decades trying to renegotiate.

We are, in our professional capacities and as Caribbean citizens, available to support the work that follows.

Respectfully,

Adrian Dunkley
President
Caribbean AI Risk Management Council

On behalf of the founding Council and members across CARICOM.

Recommendations for Immediate Action

For the Bureau of CARICOM Heads of Government: place AI on the agenda of the next Conference and designate a Lead Head.

For the CARICOM Secretariat: commission the Regional AI Risk and Capability Assessment within the current calendar year.

For national Cabinets: nominate a senior official as the AI policy focal point and ensure they are resourced to engage in CARICOM coordination.

For Caribbean regulators: align AI supervisory expectations to the NIST AI Risk Management Framework, the EU AI Act high-risk provisions, and ISO 42001, pending regional convergence.

For Caribbean parliamentarians: hold at least one parliamentary session in the next twelve months on AI policy in your jurisdiction, with civil-society and academic participation.

For Caribbean citizens, journalists, and civil society: hold us to the commitments in this letter. Public scrutiny is the most effective accelerator of regional policy that we have.

Frequently Asked Questions

Is this letter a formal communication to CARICOM?

This is an open letter, intended both as a respectful submission to the Heads of Government and as a public statement that Caribbean citizens, organisations, and journalists can reference. A formal communication will follow through the appropriate institutional channels.

Does CAIRMC speak for the Caribbean Community on AI?

CAIRMC is an independent Caribbean professional body. We do not speak for CARICOM. We speak to CARICOM, on behalf of Caribbean risk and compliance professionals and the broader community of citizens, academics, and practitioners working on responsible AI in the region. The decisions called for in this letter must be made by Heads of Government and by national institutions.

Why now rather than in two years?

The international standards that will define how AI is governed across the major economies are being set in 2026 and 2027. Caribbean input into those standards is most effective before they are finalised, not after. The cost of delay is the cost of being a rule-taker rather than a rule-shaper.

How can a Caribbean professional or citizen support this work?

Three practical steps. Read and share this letter. Ask your member of parliament where they stand on the recommendations in it. Join the CAIRMC community to participate in the regional capacity programme.

What if individual member states disagree with parts of the proposed framework?

A regional AI framework should be a floor, not a ceiling. National governments retain full sovereignty to go further on AI policy in their jurisdictions. The purpose of regional coordination is to ensure that the floor is high enough to protect Caribbean citizens and credible enough to be recognised by international partners.