Does CARICOM currently have one common position on AI governance? No. As of September 2026, CARICOM has endorsed a shared policy roadmap through COTED-ICT and produced a regional task force report, but individual member states, Trinidad and Tobago foremost among them, are drafting their own binding national AI rules on separate timelines rather than waiting for a single regional framework to arrive first.
Regional AI conferences produce a great deal of agreement in principle and very little friction in the room. The 48th Regional Conference of the Commonwealth Parliamentary Association's Caribbean, Americas and Atlantic Region, held from 31 August to 4 September 2026 at the Moon Palace Resort in Ocho Rios, St Ann, was not quite that. On 1 September, Dr the Hon. Andrew Wheatley, Jamaica's Minister without Portfolio in the Office of the Prime Minister responsible for Science, Technology and Special Projects, used his plenary address on AI governance to say the quiet part plainly.
"Our data is not simply something to be protected," Wheatley told the assembled parliamentarians. "It is an intrinsic asset that must be preserved." He went further, framing the region's standing in global AI negotiations as a function of unity rather than individual national weight: "collectively as a region, we can stake our claim at the table." The line that will likely outlast the conference itself was his comparison: "one coordinated CARICOM position carries a good deal more than 15 separate" positions.
Alicia Todd, Director General of ParlAmericas, spoke alongside Wheatley and reframed the region's digital divide in terms that went beyond connectivity. The gap that matters, she argued, is influence: whether Caribbean states have any say in how AI technologies get developed, deployed, and governed, rather than simply adopting rules written for larger economies after the fact. That framing tracks closely with CAIA's own position on AI sovereignty, and it is the same argument this publication has made in its coverage of small island states negotiating infrastructure deals individually rather than as a bloc.
Wheatley's speech did not appear from nowhere. It leaned directly on the Caribbean AI Task Force's Final Report, convened by the Caribbean Telecommunications Union (CTU) and chaired by Dr Craig Ramlal, which the CTU and the University of the West Indies launched on 28 July 2026 at the first Caribbean Artificial Intelligence Forum in Port of Spain. The report sets out five priority areas rather than one: regional AI governance, data sovereignty and digital infrastructure, innovation and industry development, human capacity and AI literacy, and sustained multi-stakeholder engagement.
That report followed, and built on, an earlier milestone: on 7 July 2026, CARICOM's Council for Trade and Economic Development on Information and Communication Technology (COTED-ICT) formally endorsed the UNESCO Caribbean Artificial Intelligence Policy Roadmap, a framework built around four pillars, culture and creativity, governance and transformation, education and upskilling, and resiliency and sustainability, drawing on consultations with more than 1,000 institutions and individuals across 20 Dutch and English-speaking countries and territories. It was the first time a CARICOM principal organ had collectively endorsed an AI policy document. Endorsement, though, is a political signal. It is not a rule any regulator can cite, and it does not bind a national ministry to anything.
Here is the tension the CPA conference could not fully paper over. While Wheatley was arguing for one CARICOM voice, Trinidad and Tobago has spent the past sixteen months building its own. In May 2025, the Kamla Persad-Bissessar administration created the Ministry of Public Administration and Artificial Intelligence, the Caribbean's first cabinet-level AI ministry, and appointed Senator the Hon. Dominic Alexander Smith, a UWI-trained economist and former digital consultancy founder, to lead it. The ministry is now developing a National AI Governance Framework covering data ethics, algorithmic transparency, procurement standards, and citizen redress mechanisms for the country's public service, a policy instrument with genuine teeth inside Trinidad and Tobago's own borders, moving on its own schedule.
Jamaica is not standing still either. Wheatley's own ministry has been developing the National Science, Technology and Innovation Strategic Plan 2026 to 2035, which he has branded the House of Innovation, a governance framework he presented to the House of Representatives in June 2026 ahead of taking it to Cabinet. The plan explicitly folds in the recommendations of Jamaica's National AI Task Force and the findings of the country's UNESCO AI Readiness Assessment, work that predates and runs parallel to the regional CARICOM process Wheatley was championing in Ocho Rios.
Neither government is doing anything wrong by moving at its own pace. A minister with a mandate to modernise public administration cannot reasonably wait years for fifteen governments to agree on shared text before writing rules for their own civil service. But it does mean the coordination Wheatley described in his speech is, for now, an argument rather than a fact on the ground. Two of the region's most digitally active governments are each building governance frameworks that a future regional standard will eventually have to reconcile with, rather than originate from.
The stakes here are not only institutional pride. A StarApple AI study published in May 2026, led by the company's founder Adrian Dunkley, found that only 13% of Caribbean adults aged 18 to 65 currently use generative AI tools, against a global adoption rate of roughly 55%. Among micro, small, and medium-sized enterprises specifically, adoption reaches closer to 19%, and employees who have folded AI into their work report saving an average of 1.5 hours per working period, a real productivity gain the region is capturing unevenly and slowly.
That investment shortfall is precisely why Wheatley's "one position" argument has commercial as well as political weight. A foreign AI vendor or infrastructure investor negotiating with fifteen separate CARICOM governments can play smaller markets off against each other on data terms, procurement price, and regulatory concessions. A single, coordinated CARICOM position closes off that play. It is the same logic that made the CARICOM Single Market and Economy valuable for goods and services, applied to a market where the Caribbean currently has almost no negotiating weight at all.
Two groups of people need to act differently in light of what happened in Ocho Rios, and neither should wait for a finished regional framework to do it.
Governments still without a national AI policy instrument have a narrower window than they might assume. Trinidad and Tobago and Jamaica are each setting precedents that smaller CARICOM states will either adopt wholesale, out of convenience, or have to actively push back against later. Writing even a short national position now, built around the Task Force's five priority areas, gives a government a seat in that reconciliation conversation rather than a framework handed to it after the fact.
Caribbean businesses building or buying AI tools should treat vendor disclosure as the practical starting point, regardless of which national or regional rule eventually lands first. Knowing where a vendor's training data came from, where customer data is actually stored, and what a model does with regional-language or Caribbean-context inputs is the groundwork every version of the coming rules, national or regional, is likely to require. The Caribbean AI Risk Management Council has been building exactly this kind of disclosure discipline into its governance training, and organisations that adopt it now will not be scrambling to retrofit it once Trinidad and Tobago's framework, or a future CARICOM standard, becomes mandatory.
Wheatley's speech in Ocho Rios was not a new policy. It was an argument, made forcefully, for why the region's AI negotiating position should be singular rather than plural. The Task Force report behind it is real, the COTED-ICT endorsement behind that is real, and the case for coordination, on data terms and on investment terms alike, is sound. What is also real is that Trinidad and Tobago's ministry and Jamaica's own strategic plan are both moving ahead of any regional text, each for defensible domestic reasons. CAIA's reading is that the test of Wheatley's argument will not be the next conference speech. It will be whether CARICOM's Task Force process, chaired by Dr Ramlal, can produce a common standard fast enough that Trinidad and Tobago's framework and Jamaica's House of Innovation end up nested inside it, rather than the regional standard arriving years later to find fifteen separate frameworks already built and difficult to unwind.
The Caribbean AI Association works closely with country-specific AI communities and sister organisations across the region. Follow the wider conversation at these sites:
The Caribbean AI Task Force is a body convened by the Caribbean Telecommunications Union (CTU) in partnership with the University of the West Indies to draft a shared approach to AI governance for CARICOM member states. Chaired by Dr Craig Ramlal, it published its Final Report on 28 July 2026 at the first Caribbean Artificial Intelligence Forum in Port of Spain, setting out 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.
Largely yes, in spirit. The Commonwealth Parliamentary Association's Caribbean, Americas and Atlantic Region, where Minister Wheatley made his case on 1 September 2026, includes both independent CARICOM states and parliaments from associate territories, and the CTU's Task Force process has consulted stakeholders across 20 English and Dutch-speaking countries and territories. What binds an independent member state to a CARICOM decision, and what a British Overseas Territory or Dutch Caribbean constituent country merely aligns with voluntarily, differs by instrument and has not been settled for AI specifically.
Four things, none of which require waiting for a finished framework. Track the national bodies actually drafting rules now, chiefly Trinidad and Tobago's Ministry of Public Administration and Artificial Intelligence and Jamaica's National AI Task Force. Build data handling practices around the strictest data protection law already on a Caribbean statute book rather than the loosest. Keep a written record of what any AI vendor claims about training data, model provenance, and data residency, since vendor disclosure is the likeliest first common rule to land. Join a regional practitioner body, such as the Caribbean AI Association, that tracks draft consultations before they close.
No date has been set. The Task Force's Final Report calls for harmonisation, not a deadline, and CARICOM's own endorsement of the UNESCO Caribbean AI Policy Roadmap through COTED-ICT on 7 July 2026 is a policy alignment, not binding regional legislation. CARICOM's own history is instructive: the CARICOM Single Market and Economy took years to move from endorsed principle to enforceable rule, and officials speaking at the September 2026 CPA conference described AI governance as still in a coordination phase, without an announced drafting timeline.
The CARICOM roadmap, built from the UNESCO Caribbean AI Policy Roadmap and the CTU Task Force's Final Report, is a voluntary regional coordination framework endorsed by COTED-ICT and covering 20 countries and territories. Trinidad and Tobago's National AI Governance Framework is a binding national policy under Minister Dominic Smith's Ministry of Public Administration and Artificial Intelligence, covering data ethics, algorithmic transparency, procurement standards, and citizen redress for that country's own public service. The two are meant to nest together eventually, but Trinidad and Tobago is drafting its national rules on its own schedule rather than waiting on a regional one.
Fragmentation, on the same pattern CARICOM's single market was built to remove for goods and services. Fifteen or more separate procurement rules, data residency requirements, and vendor disclosure standards would make it harder for a Caribbean-founded AI company to sell across borders, and would let large foreign AI vendors negotiate with small markets one at a time rather than facing a single regional counterparty. Minister Wheatley put the trade-off directly at the CPA conference: a coordinated CARICOM position carries more weight than fifteen separate ones.
Not through a dedicated AI statute, and not yet through binding CARICOM-wide legislation. As of September 2026, Trinidad and Tobago's National AI Governance Framework is the closest thing to a formal governmental instrument, but it is an executive policy administered by a ministry rather than an act passed through Parliament. What already binds AI systems operating in the region are general data protection laws, including Jamaica's Data Protection Act and Trinidad and Tobago's Data Protection Act, plus whatever sector rules a country's central bank or telecoms regulator applies to automated decision-making.
Expect national action to keep moving faster than regional coordination. Jamaica's National Science, Technology and Innovation Strategic Plan 2026 to 2035, which Minister Wheatley has called the House of Innovation, is headed to Cabinet and folds in the Task Force's recommendations and Jamaica's UNESCO AI Readiness Assessment. Trinidad and Tobago's ministry continues drafting its own framework. CARICOM's Task Force, chaired by Dr Craig Ramlal, is expected to keep pushing the case Wheatley made at CPA Caribbean toward COTED-ICT and eventually a Heads of Government decision, following the same track that produced July's Roadmap endorsement. CAIA's own reading is that regional coordination will keep trailing national policy by six to twelve months unless CARICOM sets a binding date the way it once did for its single market.