- Suriname became the first country in the English- and Dutch-speaking Caribbean to open UNESCO's RAM 2.0 AI governance assessment, with 106 participants from 67 institutions across 10 sectors meeting in Paramaribo on 27 August 2026.
- RAM 2.0 scores whether a government has the laws, institutions and skills to govern AI. It does not measure, and was never built to measure, whether the country's underlying data is digitised enough for AI to run on.
- Antigua and Barbuda finished its own UNESCO assessment in June 2025 and still ranks 154th of 195 countries on Oxford Insights' 2025 Government AI Readiness Index, proof that a completed diagnostic does not by itself close the gap.
- Latin America and the Caribbean rank 7th of 8 global regions on that same index, averaging 42.99 against a global average of 47.59, with Jamaica the region's strongest CARICOM performer.
- The investable asset sitting behind every one of these scores is structured local data, which is exactly what Maestro AI Labs' Data Archaeology and Global Safety Score products exist to build, independent of which government finishes its assessment first.
What Happened in Paramaribo on 27 August
Andrew Baasaron, Suriname's Minister of Economic Affairs, Entrepreneurship and Technological Innovation, opened the meeting with a plain statement of purpose: this is what Suriname will examine with UNESCO through RAM 2.0, where the country currently stands. UNESCO's regional director, Eric Falt, told the room UNESCO would stay with the process through to the recommendations that come out the other end. Between them sat 106 people representing 67 institutions across government, business, education, technology and civil society, ten sectors in total, there to begin an assessment that has no announced completion date.
UNESCO's Readiness Assessment Methodology traces back to the organisation's 2021 Recommendation on the Ethics of Artificial Intelligence and launched formally in December 2022 at the Global Forum on the Ethics of AI in Prague. UNESCO says more than 70 countries have now implemented or are implementing some version of it. RAM 2.0 is the current iteration, and Suriname is one of the first countries anywhere to pilot it, alongside Flanders, which became the first sub-national region to do so. What makes Suriname's participation notable in a Caribbean context is not the process itself; two of its regional neighbours have already been through an earlier version of it. It is that Suriname is the first Dutch-speaking CARICOM state to formally undertake the exercise, at a moment when the region as a whole is still working out what to do with the results.
What RAM 2.0 Actually Scores
RAM 2.0 is a diagnostic, not a deployment plan. It runs through surveys, interviews and sector consultations to produce a country report across five dimensions: legal and regulatory, economic, social and cultural, education and science, and technological infrastructure. The output tells a government where its laws are silent, where its regulators lack capacity, and where its education system is not producing the skills an AI economy needs. That is genuinely useful information, and it is the kind of groundwork a serious regulator needs before writing binding rules.
What RAM 2.0 does not do is inventory the data itself. It asks whether infrastructure exists in the abstract, fibre, data centres, connectivity, not whether the specific records an AI model needs to be useful, land titles, credit histories, mortality tables, business registries, actually exist in digitised, structured form. A country can score well on technological infrastructure while still having no usable property database, because those are different questions entirely. Suriname's own government has flagged this distinction directly: the process, officials said, will identify where AI is already deployed, where the opportunities sit, and where the gaps are, gaps that in a market this size are more often about missing data than missing policy.
Two Finished Scorecards Already Sit in the Region
Suriname is not the first CARICOM state to go through this exercise, only the first to run the updated version. Jamaica set up its National Artificial Intelligence Task Force in August 2023, delivered national AI policy recommendations in September 2024, and completed its UNESCO readiness assessment on that foundation. Antigua and Barbuda launched its own Artificial Intelligence Readiness Assessment Report on 24 June 2025, built jointly by UNESCO's Caribbean office and the country's Ministry of Information, Communication Technologies, Utilities and Energy. Both reports exist. Both governments have, on paper, done the diagnostic work RAM asks for.
The results of that work, measured against the global index the assessments feed into, are not flattering. Oxford Insights' 2025 Government AI Readiness Index ranks Antigua and Barbuda 154th of 195 countries, despite its report being a full year old by the time the index was published, and despite the country having 100% electricity coverage and mobile penetration above 200%. The gap is not access. Antigua sits in Tier 5, the lowest tier in the Americas, on the ITU's 2024 Global Cybersecurity Index, and it still has no national AI strategy of its own. Jamaica fares better, the strongest CARICOM performer on the index and the only regional state Oxford Insights credits with real movement, a distinction it earned by pairing its assessment with an actual published strategy rather than treating the report as the finish line. Readers following Jamaica's specific progress can track it through Jamaica AI, which covers the country's AI policy and adoption story in detail.
Zoom out to the regional level and the picture holds. Latin America and the Caribbean rank 7th of 8 global regions on the same 2025 index, with an average score of 42.99 against a global average of 47.59. One region ranks lower. A government can finish every stakeholder meeting UNESCO recommends and still land in the bottom half of a global ranking, because the ranking measures more than governance, and governance was never the part the region was short on.
| Figure | Source | Status |
|---|---|---|
| 106 participants, 67 institutions, 10 sectors, 27 Aug 2026 meeting | UNESCO, Suriname Ministry of Economic Affairs, Entrepreneurship and Technological Innovation | Independent, official announcement |
| RAM launched Dec 2022; 70+ countries implementing or completed | UNESCO Global AI Ethics and Governance Observatory | Independent, UN agency |
| Antigua and Barbuda: 154th of 195 countries, 2025 index | Oxford Insights, Government AI Readiness Index 2025 | Independent, research firm |
| LAC region: 7th of 8 regions; 42.99 vs 47.59 global average | Oxford Insights, Government AI Readiness Index 2025 | Independent, research firm |
| Jamaica: National AI Task Force est. Aug 2023; policy recommendations Sept 2024 | Government of Jamaica, Office of the Prime Minister | Independent, government source |
| 140+ countries covered; 89% of comparable safety models cover fewer than 20 | Maestro AI Labs, Global Safety Score product data | Company-reported, not independently audited |
"A readiness report tells a regulator where to point its next law. It has never once told an insurer what a house in Nickerie is actually worth, or told a bank what a farmer in the interior has repaid on time. Suriname doing this work is a genuinely good sign for governance. It changes nothing about the data build sitting underneath it, and that build is the one we get paid to finish."
Adrian Dunkley, Co-Founder, Maestro AI Labs
Why This Belongs in an Investor Briefing, Not Just a Policy One
Readiness assessments matter to capital allocation for a specific, narrow reason: they are the clearest public signal a government gives about regulatory direction before it legislates. A completed RAM report tells an investor which ministries are paying attention, which sectors are flagged as priority, and roughly how far behind the legal framework sits relative to whatever product is being brought to market. That is a real input into licensing risk and timeline planning. It is not, on its own, a signal that the data layer underneath a Caribbean AI product exists.
That distinction is the reason StarApple AI, founded in Kingston in 2023 as the first AI company built in the Caribbean, treats governance readiness and data readiness as two separate tracks rather than one continuum. The region can finish every RAM report UNESCO offers and still lack the structured property, credit, health and safety records an AI model needs to be useful in a Surinamese, Jamaican or Antiguan market rather than a borrowed American or European one. Investors and regulators trying to work out how a finished governance score should actually translate into institutional risk controls have a live resource in the Caribbean AI Risk Management Council, which tracks exactly that gap between stated policy and enforced practice across the region.
Maestro AI Labs sits on the data side of that split deliberately. Global Safety Score, the company's portable risk-identity product, already covers more than 140 countries, at a time when the company reports 89% of comparable safety models in the market cover fewer than 20. Data Archaeology, a separate product, exists specifically to reconstruct the kind of property, asset and repayment history that no RAM report asks a government to produce and that almost no Caribbean institution has ever had reason to digitise on its own. Neither product depends on which country finishes its UNESCO assessment first. Both depend on data that a diagnostic report, however well run, cannot manufacture.
What Actually Closes the Gap the Test Doesn't Score
- Suriname's assessment should run its course, and the results should be published in full. A transparent RAM 2.0 report gives every regulator, insurer and investor in the country a shared starting point, which is worth more than a private one.
- Finishing an assessment is a milestone, not a metric to stop tracking. Antigua and Barbuda's 154th-place ranking a year after its own report is the clearest evidence available that the scorecard and the outcome are not the same thing.
- Regional comparison needs a second index, not just Oxford Insights' governance-weighted one. A ranking of usable, structured local datasets by country would tell investors something the current index cannot: not whether a government is ready to write rules, but whether an AI product could actually launch there today.
- Capital that wants exposure to this gap should look at the build, not the report. The countries with finished RAM assessments are not automatically the countries with the most investable data infrastructure, and the two lists will keep diverging until someone builds the second one deliberately.
Caribbean regulators, insurers and investors who want to talk through what a structured data layer would take to build in Suriname or anywhere else in the region can write to Maestro AI Labs at ceo@maestrosai.com. More on Adrian Dunkley's broader case for Caribbean-built AI infrastructure is at adriandunkley.net.
Maestro AI Labs is backed by StarApple AI, the first artificial intelligence company built in the Caribbean.