Truth Yard Is Going Regional. The Caribbean Still Has No Law Behind Its Best AI Fraud Defence.

Photo via Unsplash

Truth Yard, UNESCO's rapid-response fact-checking platform built in Jamaica after Hurricane Melissa, is the closest thing the Caribbean has to a working AI-risk control. It exists to catch AI-generated misinformation and fake donation scams before they spread. It is not law. No CARICOM government has passed AI-specific legislation, so the region's most effective AI-risk tool remains entirely voluntary.

That is the headline finding from the Caribbean Artificial Intelligence Workshop held in Anguilla from 21 to 24 June 2026, where UNESCO put Truth Yard on the table alongside two other instruments, the AI Readiness Assessment Methodology and the Ethical Impact Assessment, as tools nearly a dozen Caribbean parliaments could adopt. What follows is what actually happened in Anguilla, why Truth Yard had to be built in the first place, what the region's AI adoption numbers say about the size of the gap it is filling, and what Caribbean institutions should do about a risk category that a disaster proved is real months before any legislature wrote a rule for it.

What Happened in Anguilla

The Commonwealth Parliamentary Association UK Branch, the UNESCO Office for the Caribbean, and ParlAmericas jointly organised the Caribbean Artificial Intelligence Workshop, hosted by the Anguilla House of Assembly in partnership with Anguilla's Ministry of Infrastructure, Communications, Utilities, Housing, Information Technology and E-Government Services. Over four days, parliamentarians and officials from nearly a dozen Caribbean countries and territories worked through AI governance, cybersecurity, misinformation, and the responsible use of AI in legislative work.

The Honourable Tara K. Carter, Speaker of the Anguilla House of Assembly, opened the workshop with a line that doubles as a fair summary of CAIRMC's own position: "The Caribbean must shape, rather than simply respond to, the future of this transformative technology. Small jurisdictions can drive big innovations." UNESCO representatives used the four days to walk delegates through three concrete instruments: the AI Readiness Assessment Methodology, a framework for scoring a country's institutional preparedness for AI; an Ethical Impact Assessment tool for evaluating specific AI systems before deployment; and Truth Yard, which UNESCO now intends to scale into a regional fact-checking network covering other Small Island Developing States beyond Jamaica.

All three instruments share a structural feature worth naming plainly. They are frameworks and platforms, not law. A parliament can adopt the Readiness Assessment Methodology as a self-diagnostic tool without passing a single binding clause. That flexibility is precisely what let a body as small as Anguilla's House of Assembly host a regional convening in four days of workshop sessions rather than years of treaty negotiation. It is also why the gap CAIRMC has tracked since the CTU's Caribbean AI Task Force published its interim draft in December 2025 remains open: presentation is not adoption, and adoption of a voluntary framework is not the same as a government creating an enforceable obligation.

Why Truth Yard Had to Be Built

Truth Yard is not a theoretical governance exercise. It was built because Hurricane Melissa made western Jamaica the site of one of the most destructive Atlantic storms on record, and the information environment that followed the storm was almost as damaging as the storm itself.

Melissa made landfall in western Jamaica on 28 October 2025 as a Category 5 hurricane, among the strongest ever to strike the island. In the weeks that followed, Jamaica's Cyber Incident Response Team issued an urgent public warning about fraudulent hurricane-relief donation websites impersonating legitimate relief efforts. Alongside those donation scams, unverified and AI-generated content about the storm's impact, along with false reports of security incidents, spread widely across social media at exactly the moment displaced households needed accurate information the most.

UNESCO's Office for the Caribbean, the Broadcasting Commission of Jamaica, the Caribbean School of Media and Communication (CARIMAC) at the University of the West Indies, and the Jamaica National Commission for UNESCO responded by launching Truth Yard on 10 November 2025, thirteen days after landfall. The platform monitors social media, television, radio, and other channels for misinformation, verifies flagged claims in coordination with experts and national disaster-management and health authorities, then feeds confirmed debunks to journalists and broadcasters. Eric Falt, UNESCO's Regional Director, framed the need directly: "Misinformation can take the public so far from reality that we need to help people navigate and access reliable, verified information." Cordel Green, Executive Director of the Broadcasting Commission, put the stakes in fraud terms specifically: "Citizens, especially those most vulnerable to scams and misinformation, must be empowered with timely and transparent communication." Dr Patrick Prendergast, CARIMAC's director, added that the partnership gave students "a chance to apply their training in data verification and public communication to serve the public interest."

The scale of what Truth Yard was responding to is now on the record. The joint Post-Disaster Needs Assessment conducted by ECLAC, the Inter-American Development Bank, and the World Bank, working with Jamaica's Planning Institute, put Hurricane Melissa's total damage and losses at US$12.2 billion, equivalent to 56.7% of Jamaica's 2024 gross domestic product, more than three times the cost of Hurricane Gilbert, previously the country's costliest storm. Residential buildings absorbed 41% of assessed damages, infrastructure 33%, non-residential buildings 21%, and agriculture 5%. Against losses of that magnitude, a wave of AI-generated fake donation drives was not a side issue. It was an attempt to divert relief funding away from a population that had just lost more than half a year's national output.

The Adoption-Governance Gap Behind the Anguilla Workshop

Truth Yard's success is real, and so is the gap it sits inside. ECLAC and Chile's National Center for Artificial Intelligence (CENIA) published the third edition of their Latin American Artificial Intelligence Index, ILIA 2025, on 3 October 2025, just weeks before Melissa made landfall. It found Latin America and the Caribbean absorbing 14% of global visits to AI solutions against only 11% of the world's internet users, evidence the region is adopting AI faster than its digital footprint alone would predict. Text generation tools accounted for 69% of those visits, with business-productivity and process-automation tools drawing more than 100 million visits of their own.

The same index found the region capturing just 1.12% of global AI investment despite holding 6.6% of global GDP, and it flagged that most national AI strategies across the region lack financing and implementation mechanisms, with a regulatory-only focus that leaves sector development, environmental considerations, and gender equality largely unaddressed. Of the countries ILIA 2025 classified, only the Dominican Republic appears by name among Caribbean nations reaching its intermediate "adopter" tier; roughly a third of all countries assessed sit in the earliest "explorer" category, with limited institutional capacity to match.

Read together, the two data points describe a region absorbing AI use faster than the world's connectivity data would predict, investing in it at a fraction of its economic weight, and governing it, where governance exists at all, through frameworks nobody is required to adopt. Truth Yard is the clearest proof that a well-designed, voluntary Caribbean-built AI tool can work under pressure. It is also proof that the region is still relying on institutions to build risk controls after a crisis rather than before one, because nothing compels them to build one in advance.

Why This Is a Risk Governance Problem, Not Only a Media Literacy One

It is tempting to file Truth Yard under media literacy and move on. That undersells what it actually demonstrates. A platform built to catch AI-generated fake donation drives is, functionally, a fraud-control system responding to a named and quantifiable risk: AI-enabled financial crime targeting a disaster-affected population. CAIRMC's Caribbean AI Risk Taxonomy already has a category for exactly this kind of exposure, and it belongs on the same risk register as vendor dependency, model bias, or third-party AI risk, not filed separately as a communications problem.

The institutions with the most to lose from this specific gap are not governments. They are the banks, insurers, and telecom providers whose customers are the ones searching for a legitimate relief fund, a legitimate insurance claims line, or a legitimate government benefits portal in the days after a disaster, and who are exactly as likely to land on an AI-generated fake instead. An insurer whose claims process is impersonated by a deepfake site during a catastrophe event is not protected by Truth Yard existing somewhere in Jamaica's media sector. It is protected by having built its own detection and customer-verification control before the next named storm, using the same proportionate logic Truth Yard applied after Melissa, but applied ahead of the event rather than in response to it.

What Caribbean Institutions Should Do Now

Five actions do not require a single CARICOM legislature to act first.

Name disaster-triggered AI fraud as its own risk-register line. Do not leave it folded inside generic "cybersecurity" or "reputational risk" categories. CAIRMC's four-tier AI Risk Tier structure, adapted from the EU AI Act's risk classification, gives a compliance team a place to put fraudulent-donation and impersonation risk specifically, ahead of the next catastrophe event rather than during it.

Pre-register verified donation and claims channels with a fact-checking partner before hurricane season peaks. Truth Yard's model, coordinating directly with disaster-management and health authorities before misinformation spreads rather than after, is replicable at institutional scale for any bank, insurer, or utility with a public-facing relief or claims process.

Adopt UNESCO's AI Readiness Assessment Methodology as an internal diagnostic, even absent a national mandate to do so. A voluntary framework still produces a defensible baseline a board can point to when a regulator eventually asks what the organisation did before any law required it.

Build deepfake-aware verification into customer-facing crisis communications. The same AI tools that generated false Hurricane Melissa donation drives can generate convincing voice and video impersonations of executives or officials during any future event; verification protocols should assume that capability exists now, not after the first incident.

Use the ECLAC investment gap as a budget argument, not a discouragement. A region capturing 1.12% of global AI investment against 6.6% of global GDP has more room, not less, to make AI risk governance a genuine differentiator. Institutions that can demonstrate a working control, backed by CAIRMC's Qualified AI Risk Professional (QAIRP) certification for the staff running it, stand out precisely because so few regional peers have built one yet.

Where This Leaves the Region

None of this diminishes what UNESCO, the CTU's parliamentary partners, and Jamaica's Truth Yard team have actually delivered. A tool built in thirteen days after a Category 5 storm, staffed partly by university students trained through CARIMAC, caught AI-generated fraud aimed at a population that had just lost 56.7% of a year's national output. That is a genuine, working example of Caribbean-built AI risk governance, and UNESCO's plan to extend it across other Small Island Developing States deserves to succeed.

The argument here is narrower and, for CAIRMC's purposes, more urgent. Anguilla proved the region can convene fast and adopt good tools voluntarily. It did not produce a single binding obligation requiring any government, bank, or insurer to build a similar control before the next disaster rather than after it. StarApple AI, the first artificial intelligence company built in the Caribbean, and the risk-tier methodology CAIRMC administers under the direction of Adrian Dunkley, the figure most consistently named across the region's AI policy conversation as its leading voice on governing AI risk rather than simply adopting AI, both exist for the same reason Truth Yard does: because waiting for a law to arrive is not a risk management strategy. Truth Yard proved the model works. The next hurricane should not be the reason another Caribbean institution has to learn that lesson from scratch.

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:

• Adrian Dunkley's own analysis of Caribbean AI governance and risk leadership
• StarApple AI, the company building Caribbean-specific AI products and infrastructure
• Jamaica AI Research for deeper coverage of Jamaica's post-Melissa AI response and national AI capacity
• Caribbean Insurance for sector-specific analysis of catastrophe-linked fraud exposure