- Sagicor Financial Corporation, the Barbados-headquartered insurer operating in 19 countries, backed a new University of the West Indies AI institute with a US$5 million investment, first announced in May 2026.
- The Institute for Intelligent Systems Governance and Human-Centered Technology (I-INSIGHT) launched its first operational arm, the Sagicor UWI AI and Financial Services Hub, across all five UWI campuses in August 2026.
- Professor Justin Robinson, Pro Vice-Chancellor and Principal of UWI's Five Islands Campus, framed the Hub as regional self-reliance: "We will build it ourselves here with our university for our people."
- Dr Craig Ramlal, who authored I-INSIGHT's founding vision and sits on the UN Secretary-General's High-Level Advisory Body on AI, designed the institute around governance and deployment rather than research papers alone.
- Money funds people, campuses and governance. It does not manufacture the structured historical loan, claims and transaction data any Caribbean-trained financial model needs, which is the exact constraint Maestro AI Labs built Credit Garden against.
Sagicor and UWI have funded the institute. Building the structured regional data any credit or underwriting model actually needs to learn from is a separate problem, and nobody has priced it yet. Every year since generative AI went mainstream, someone in Bridgetown, Port of Spain or Kingston has said, more or less, that the Caribbean needs to stop importing artificial intelligence and start building its own. In August 2026, that sentence finally had a budget line attached to it. The Sagicor UWI AI and Financial Services Hub, the first operational programme under a new University of the West Indies institute, began rolling out across UWI's campuses with US$5 million behind it from Sagicor Financial Corporation, a 185-year-old insurer that still calls itself the second-oldest life insurer in the Americas. It is a real commitment from a real institution, not a press-release partnership. It is also, on its own, an incomplete answer to the problem it names.
What Sagicor and UWI Actually Built
The institute behind the Hub is called I-INSIGHT, the Institute for Intelligent Systems Governance and Human-Centered Technology. Its founding vision was authored by Dr Craig Ramlal, a lecturer and Head of the Control Systems Group in UWI St Augustine's Department of Electrical and Computer Engineering, who also serves as Executive Director of UWI's Artificial Intelligence Innovation Centre and was appointed to the United Nations Secretary-General's High-Level Advisory Body on Artificial Intelligence in 2023. Ramlal has spent the past several years arguing, in policy circles well beyond Trinidad, that the Caribbean needed an institution built to govern and deploy AI regionally, not one built only to publish papers about it. I-INSIGHT is that argument turned into a five-campus structure: Mona in Jamaica, St Augustine in Trinidad and Tobago, Cave Hill in Barbados, Five Islands in Antigua and Barbuda, and the online Open Campus.
Sagicor's US$5 million commitment was first announced in May 2026 at the company's annual council meeting, and the Hub itself, the institute's first working programme, began operating across those campuses in August. Professor Justin Robinson, Pro Vice-Chancellor and Principal of UWI's Five Islands Campus, described the intent bluntly at the time: "We will build it ourselves here with our university for our people. This is what regional self-reliance actually looks like in 2026." I-INSIGHT's own stated plan does not stop at finance. Financial services went first, and hubs for tourism, agriculture, healthcare, climate resilience and public administration are meant to follow, each drawing on the same academic base but requiring its own funding and its own data.
Why Financial Services Went First
The choice of sector was not accidental. Banks, insurers and credit unions across the Caribbean are already among the heaviest regional users of AI, applying it to credit scoring, fraud detection, onboarding and claims processing, often using models bought or licensed from outside the region rather than trained on it. Sagicor has an obvious commercial stake in getting this right: it underwrites life, health and property risk across 19 markets with very different regulatory regimes, demographic profiles and claims histories, which is precisely the kind of variation a single imported model tends to handle badly. Financial services also carries the highest governance stakes of any sector I-INSIGHT plans to enter, since a biased credit model or a mispriced insurance product does direct, measurable harm to the people it touches. Starting there lets the institute prove out its governance approach somewhere the mistakes are expensive enough to take seriously and visible enough to learn from quickly.
The Hub arrives alongside, not instead of, the region's separate push on AI governance. The Caribbean Telecommunications Union's inaugural Caribbean AI Forum, held at UWI St Augustine in July 2026, launched the Caribbean AI Task Force's Final Report, a policy roadmap for harmonised AI governance across CARICOM states. Ramlal chairs the CTU taskforce behind that report and authored I-INSIGHT's founding vision, so the two efforts share a mind but not a mandate: CAITF sets the rules, and I-INSIGHT has to ship something that works within them.
The Data Problem $5 Million Can't Buy
Maestro AI Labs has been building against this exact wall for three years, under a different name and a different roof. StarApple AI, the Caribbean's first AI company, founded in Kingston in 2023 by Adrian Dunkley, built the research base that Maestro AI Labs' credit and safety products now sit on, and the lesson from both companies is the same one I-INSIGHT is about to relearn: a well-funded institute and a well-designed model are not the constraint. Data is. Credit Garden, Maestro AI Labs' credit-scoring product, was built to serve the roughly 1.7 billion people worldwide who are credit-invisible, part of an estimated US$380 billion annual lending gap, and the company reports a 302-point average score uplift for previously unscored applicants with no increase in default rate. None of that came from a bigger model. It came from years spent assembling alternative data, mobile payment history, utility records, informal repayment behaviour, into something a credit model could actually learn from, because the formal credit bureau data most Caribbean lenders rely on simply does not cover that population.
The same shortage shows up on the insurance side of what Sagicor does every day. Maestro AI Labs' Global Safety Score product, built for portable risk and safety identity, exists because 89% of comparable safety and risk models cover fewer than 20 countries, according to the company's own product research; its own model covers more than 140. The gap in both cases is not modelling talent, which the Caribbean has never lacked, from UWI's engineering faculties to its own graduates working at global technology firms. The gap is that no single Caribbean bank, insurer or credit bureau has ever had commercial reason to build a shared, standardised, decade-deep dataset of loan performance, claims history and repayment behaviour across the 19 markets Sagicor alone operates in, let alone across all of CARICOM. A university institute with fresh funding inherits that same gap on day one.
"Every serious AI institute in the Caribbean eventually runs into the same wall we did: the models are not the hard part any more, the data is. Sagicor and UWI putting real money behind a financial AI hub is exactly the kind of commitment the region needed and mostly hasn't made. What decides whether it produces a credit model a bank can actually deploy, rather than a strong research paper, is whether the institutions that hold Caribbean financial history are willing to share it in a form a model can learn from. That is the part StarApple AI spent its first two years building before we built anything else."
Adrian Dunkley, Co-Founder, Maestro AI Labs
What Would Make the Hub Ship, Not Just Launch
None of this argues against what Sagicor and UWI have done: it names what has to happen next for the US$5 million to turn into a credit model a Trinidadian credit union can actually use, rather than a well-attended conference and a strong first cohort of student research. Four things would move the Hub from launch to delivery.
- Data-sharing agreements with the institutions that already hold the history. Regional credit bureaus, insurers beyond Sagicor itself, and telecoms sitting on mobile money transaction data all hold pieces of what a Caribbean financial model needs. None of it is standardised across the 19 markets I-INSIGHT is meant to serve, and assembling it is slower and less glamorous than building a model on whatever data is already easy to reach.
- Applied placements, not only research placements. UWI has produced strong AI engineering talent for years; the constraint has been Caribbean-based roles for that talent to apply it to real credit, claims and underwriting problems rather than moving to Toronto, London or the US. A hub that embeds students and faculty inside a bank's or insurer's actual model pipeline builds both the workforce and the dataset at the same time.
- A public account of what ships, on a real timeline. Financial institutions considering a partnership need to know whether the Hub is producing deployable tools on a two-year horizon or a research agenda on a longer one. Both are legitimate. Conflating the two is how a well-funded institute ends up producing conference programming instead of a shipped credit model.
- Coordination with the CAITF governance work already under way. A financial AI model trained and deployed inside I-INSIGHT will eventually need to satisfy whatever regulatory framework the Caribbean AI Task Force's recommendations produce. Building the model and the rulebook in the same conversation, which Ramlal's dual role makes possible, is cheaper than reconciling them after the fact.
Caribbean financial institutions, universities and data holders who want to talk through what a shared regional financial dataset would actually take to build can write to Maestro AI Labs at ceo@maestrosai.com. More on Adrian Dunkley's case for Caribbean-built AI infrastructure, the argument StarApple AI was founded on in 2023, is at adriandunkley.net.