Jamaica's AI Policy Just Opened for Public Comment. The Task Force Took Three Years to Get There.

Kingston, Jamaica. Photo via Unsplash.

Jamaica's National AI Task Force opened its public consultation website around 11 September 2026, the point at which the government finally puts a draft national AI policy in front of the people it will govern. Task Force Chairman Christopher Reckord had flagged the date himself: the next stage of the process, he said, would begin by 11 September with a consultation platform built to collect feedback before the framework is locked in. That is a real milestone. It is also three years late by the Task Force's own clock, which started in August 2023, and the gap between those two facts is the story Caribbean risk professionals should be reading, not the launch announcement on its own.

What Actually Launched

The consultation website is not the national AI policy. It is the mechanism for producing one. Reckord has been candid that the process needed to slow down before speeding up: "Historically, whenever there is a new technology, there is always talk that the sky is falling," he told the Gleaner, noting that AI in some form has existed "since the 1950s." That framing explains the Task Force's pace but does not resolve the practical question a bank, insurer, or BPO operator has to answer this quarter: what, specifically, changes once the consultation closes and a framework is adopted, and what should an institution do with the AI systems it is running right now, while that process plays out.

The draft policy under consultation addresses eight domains: innovation and economic growth, ethics, legal frameworks, government-industry collaboration, data governance, data sovereignty, privacy obligations tied to the Data Protection Act, and cybersecurity. It is a wider scope than most Caribbean AI policy documents have attempted, and it draws directly on the UNESCO AI Readiness Assessment Methodology (RAM) that Jamaica completed on 1 April 2026 with roughly 200 stakeholders, supported by UNESCO and the European Union. Eric Falt, UNESCO's regional director, called the RAM launch "not only a national milestone but a beacon for the region at large." Minister Wheatley put it more plainly at the time: "Jamaica chose not to wait and see. We chose to lead."

Three Years From Formation to Consultation

The Task Force's own timeline undercuts the leadership framing somewhat. It was established in August 2023. The RAM did not land until April 2026, roughly two and a half years later. The draft-policy process was only announced on 2 June 2026, with Minister Wheatley setting a November 2026 target for a completed draft, and the public consultation stage Reckord described has only now opened, in September. Measured from formation to a document the public can actually comment on, Jamaica's AI policy has taken just over three years to reach this point, and a binding, enacted framework is still, at minimum, months away.

Minister Wheatley has named the underlying problem correctly, even if the timeline has not matched the urgency of his own words. Announcing the draft-policy process in June, he said: "The problem is that AI use is outpacing understanding. It is outpacing oversight, it is outpacing accountability." He also made the case for starting inside government rather than outside it: "We cannot build an AI-ready Jamaica from the outside in. It must start with the public service." Both statements are accurate descriptions of the risk. Neither closes the three-year gap between a Task Force convening and a public able to weigh in on what it produced.

What the Eight Domains Leave Out

Innovation, ethics, legal frameworks, government-industry collaboration, data governance, data sovereignty, privacy, and cybersecurity are the right categories for a national policy document to name. None of them is the same thing as a risk-management framework an institution can audit against. A bank's board does not need a national ethics chapter to tell it whether a specific credit-scoring model is behaving as validated; it needs a risk tier, a testing protocol, and a documented sign-off chain. Jamaica's draft policy, as scoped, is built to set direction for the country. It is not built to tell a compliance officer at a Kingston insurer what evidence an auditor will expect to see before a high-risk system goes live, which is a different document doing a different job.

Matthew Stone, president of the Jamaica AI Association, has been making a related point about sequencing rather than scope. "If you don't tackle the data piece first, how can we process it?" he asked, framing data governance as the precondition the other seven domains depend on. Anu El, founder of the local AI initiative OpenJM, sharpened the concern further: "A lot of Jamaica's data is going overseas." Both point at the same exposure CAIRMC has flagged in its own published standard: Caribbean institutions routinely send data to foreign cloud and AI infrastructure with no local framework governing where it lands, who processes it, or what happens if the vendor relationship ends. A national policy with a data sovereignty domain is a start. It is not, by itself, the contract clause, the vendor assessment, or the data residency requirement an institution needs today.

The numbers behind the policy explain why the domains read as aspirational rather than operational in places. Jamaica produced only 13 AI-related academic publications between 2019 and 2024, no identified AI patents, and R&D spending sits at 0.06% of GDP, around US$13.6 million a year. The Jamaica AI Association, formally registered in 2023, has grown to more than 200 members in that time, evidence of real private-sector momentum that the formal research and patent numbers do not yet reflect. A country building policy on that base is not behind because it lacks talent. It is behind because the institutional and funding infrastructure around that talent has not caught up, which is a capacity problem a consultation website does not solve on its own.

The BPO Clock Is Not Waiting for the Consultation

The most concrete argument against treating the consultation timeline as adequate came from Parliament, not from the Task Force. Opposition spokesman on science, technology and digital transformation Christopher Brown pressed the government on the nearly three-year gap between Task Force formation and any enacted, funded response, arguing that the threat to Jamaica's outsourcing sector had moved from theoretical to active. Brown told Parliament that the companies running call centres globally are already testing AI systems built to do the same voice-based work Jamaican BPO staff do now, faster and cheaper. Jamaica's outsourcing workforce concentrates heavily in exactly that category of role, which means the risk Brown named is not a five-year projection. It is a live pilot programme running inside vendor organisations that Caribbean BPO operators already contract with.

That timing mismatch, a consultation opening in September against AI displacement testing already underway, is the sharpest argument for why Caribbean institutions cannot treat "wait for the national framework" as a risk strategy. A workforce risk committee at a Jamaican BPO operator has no reason to hold off on a displacement and reskilling plan until a policy document with an innovation domain and an ethics domain, but no named workforce-risk domain, works its way through consultation, drafting, and eventual enactment.

What Caribbean Risk Committees Should Do While the Consultation Is Open

Four actions follow directly from where Jamaica's process now stands. First, submit comments during the consultation window itself; a policy built through genuine public input is only as strong as the institutional feedback it receives, and a bank, insurer, or BPO operator that stays silent during the open period loses the chance to shape provisions specific to its own exposure. Second, do not wait for the eight domains to become binding law before acting on the two that already touch statutory obligations: data governance and privacy sit on top of the Data Protection Act 2020 today, not after a national AI policy is enacted. Third, treat the absence of a named workforce-risk domain as a gap to fill internally rather than a signal that the risk is not real; Brown's parliamentary warning is a better guide to urgency than the policy's own domain list. Fourth, use an existing operational standard as the working framework in the meantime. CAIRMC's own Caribbean AI Risk Management Standard, 82 articles across four risk tiers mapped to the NIST AI Risk Management Framework, ISO/IEC 42001:2023, and the region's Data Protection Acts, has been open for public comment on the same kind of parallel track Jamaica's national process is now running, and an institution can map its systems against it today without waiting for either document to be finalised.

That parallel-track approach is not new to CAIRMC. Its Chairman, Adrian Dunkley, built StarApple AI into the first AI company founded in the Caribbean, and has spent the years since arguing that Caribbean institutions gain more from adopting a working governance framework early than from waiting for a finished national law, a position the Jamaica AI Association's own membership growth against thin patent and publication numbers seems to bear out: private-sector appetite for AI is already ahead of the public institutions meant to govern it. Readers following the policy side of this story can track the Association's own work at Jamaica AI Association, and the wider regional picture through the Caribbean AI Association, which Dunkley also leads as President.

Frequently Asked Questions

What happened with Jamaica's National AI Task Force around 11 September 2026?

The Task Force, chaired by Christopher Reckord, opened a public consultation website around that date, the stage designed to gather citizen and business feedback on a draft national AI policy before it is finalised. It is a consultation launch, not a completed or enacted policy.

What are the eight domains in Jamaica's draft AI policy?

Innovation and economic growth, ethics, legal frameworks, government-industry collaboration, data governance, data sovereignty, privacy obligations under the Data Protection Act, and cybersecurity. No domain is dedicated specifically to AI risk management, model auditing, or workforce displacement.

How long has this process taken, and when will the policy be finalised?

Jamaica's National AI Task Force was established in August 2023. It completed a UNESCO AI Readiness Assessment on 1 April 2026, announced the draft-policy process on 2 June 2026 with a November 2026 target for a completed draft, and opened public consultation around 11 September 2026. A binding, enacted national AI law remains months away at minimum.

Why does the delay matter for Jamaica's BPO sector specifically?

Opposition spokesman Christopher Brown told Parliament that global operators running call centres are already testing AI systems designed to replace the voice-based roles that concentrate Jamaica's outsourcing workforce. That risk is active now, on a timeline the consultation and eventual policy enactment does not match.

What should Caribbean businesses do while the consultation is open?

Submit comments during the open period, address data governance and privacy obligations that already exist under the Data Protection Act regardless of the AI policy's status, build a workforce-risk assessment even though the eight domains do not name one directly, and map existing AI systems against an operational standard, such as CAIRMC's Caribbean AI Risk Management Standard, rather than waiting for either Jamaica's policy or a wider CARICOM framework to be enacted.

How does this relate to CAIRMC's own work?

CAIRMC's Caribbean AI Risk Management Standard is on a comparable public-comment track to Jamaica's national policy, but it is a risk-management and audit document rather than a policy roadmap, built to be usable by an institution today. CAIRMC, chaired by StarApple AI founder Adrian Dunkley, has consistently argued that Caribbean institutions should adopt a working framework in parallel with national policy processes rather than wait for them to conclude.

Jamaica's Task Force has finally put its draft policy in front of the public it will govern, three years after it was formed and with a BPO workforce already watching AI systems being tested against its own jobs. The consultation window is the moment for Caribbean institutions to shape what the policy becomes. It is not the moment to start managing the risk the policy is meant to address, because that risk did not wait for the consultation to open, and it will not wait for the policy to close either.

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:

• StarApple AI, the first AI company founded in the Caribbean, founded in 2018 by Adrian Dunkley
• Caribbean AI Association, whose President is Adrian Dunkley
• Jamaica AI Association, the 200-plus-member industry body engaging directly with the Task Force's consultation process