Caribbean AI Adoption Is Outrunning New Governance Rules

John Ramage, Governor George Clinton, ca. 1785. The Metropolitan Museum of Art, public domain.

Caribbean legislators met as a region on AI governance for the first time in June 2026, in Anguilla. By then most of the region's startups were already using AI tools every day.

As of August 2026 no Caribbean government has passed a dedicated AI law. Businesses adopted first, and the rules are still being drafted. This piece covers what the workshop produced, how far ahead adoption is, why the gap affects investment, and what a business should put in place now.

What the Anguilla workshop produced

The Anguilla House of Assembly hosted the Caribbean AI Workshop from 21 to 24 June 2026. It was organised by the Commonwealth Parliamentary Association UK Branch, the UNESCO Office for the Caribbean and ParlAmericas, with Anguilla's Ministry of Infrastructure, Communications, Utilities, Housing, Information Technology and E-Government Services. UNESCO presented its AI Readiness Assessment Methodology and its Truth Yard fact-checking initiative. Delegates finished with an outcome document committing their legislatures to build ethical, human-centred AI governance frameworks.

It matters that a small territory hosted it. Small states usually receive technology rules written elsewhere, and here a dozen governments sat down to coordinate their own.

An outcome document is a commitment, not a statute. Moving from agreed principles to a bill that binds a bank, a hospital or a ministry is where regional policy has usually stalled, and nothing yet shows this one will move faster.

How far ahead business adoption already is

In May 2026 the Inter-American Development Bank published Startups x AI: An Overview of Artificial Intelligence Adoption in Latin America and the Caribbean. It surveyed startups that use AI as a tool inside an existing product or process, not companies built around AI models. Generative AI (tools that write text or make images) was used by 85% and predictive AI (models that forecast from past data) by 75%, mostly in marketing, product development and decision-making.

The report covers the whole region and says its Caribbean sample is thinner than its Latin American one, so read the numbers as a direction for the Caribbean. On the ground it looks like an accounting firm drafting client letters with a chatbot, a Kingston logistics company feeding shipment data into a route model, and an agency writing first-draft ad copy with AI. None of that needed a regulator's permission, because no regulator had issued anything specific.

Nineteen months from roadmap to workshop

UNESCO published the Caribbean Artificial Intelligence Policy Roadmap in November 2024. It drew on more than 1,000 institutions and individuals in 20 English- and Dutch-speaking Caribbean countries across eight regional meetings, and it rests on four pillars: culture and creativity, governance and transformation, upskilling and education, and resilience and sustainability.

Nineteen months passed before legislators met to work out how to apply it. That is not slow by the standards of government consultation. In the same period AI vendors released new model versions every few months, and businesses adopted each one. Legislation runs in years and products in months, and no region has solved that mismatch. What separates regions is how quickly they narrow the gap once it is visible.

Why unclear rules keep investment low

The ECLAC-CENIA Latin American Artificial Intelligence Index, released in October 2025, found that Latin America and the Caribbean produce 6.6% of global GDP but attract only 1.12% of global AI investment. No country in the region reaches the world average for AI investment relative to GDP per capita.

Regulatory uncertainty is one reason. An investor can model risk in a market with a written, predictable AI framework. Fifteen CARICOM members that might each pass different rules, or none, are harder to price, so capital goes elsewhere. Talent, market size and exits matter too, and I cannot say how much of the gap a shared framework would close on its own.

Why waiting for the law is the wrong default

Across Caribbean technology waves, businesses have moved first and institutions have caught up once the pressure became public. AI is following that order at a faster pace. A business that waits for a statute before touching AI does not avoid the eventual rules. It meets them later, with less experience, behind competitors who already use the tools.

The larger risk in the region is thousands of untrained staff making AI decisions without ever having seen a responsible workflow. That is why we run free AI literacy training through AI Jamaica, and why the steps below start with records and people.

What a Caribbean business owner should do this quarter

1. The owner or general manager: list every AI tool staff use, including free chatbots, in a shared spreadsheet with what data goes into each. One afternoon is enough for a firm of under 50 people.
2. Whoever handles customer data: for each tool that receives personal data, check where the vendor stores it and whether your data protection law allows that transfer. In Jamaica, read the Office of the Information Commissioner's guidance on the Data Protection Act.
3. Department heads: name the decisions that affect a customer's money, health or legal standing, and require a person to approve any AI output in those areas before it goes out. Write the rule down in one page.
4. Procurement or the owner: before signing with an AI vendor, ask in writing how the system reaches its outputs, whether your data trains its models, and how to delete it. Drop vendors who will not answer.
5. One named staff member: spend an hour a month tracking AI rules in your sector and country, starting with your regulator's website and your parliament's order paper, and report changes to management.