CARICOM Just Created a Blue-Ribbon AI Commission. Trinidad's Deepfake Fraud Wave Shows Exactly Why It Could Not Wait

TL;DR

• CARICOM Heads of Government, meeting in Gros Islet, Saint Lucia on 3 August 2026, agreed to establish a Blue-Ribbon Commission on Artificial Intelligence to advise on AI policy, governance standards and capacity building.
• The Commission follows CARICOM's first collective endorsement of an AI policy document: COTED-ICT adopted the UNESCO Caribbean AI Policy Roadmap on 7 July 2026, built from eight consultations across 20 territories and more than 1,000 institutions.
• Neither document names the threat already documented in Trinidad and Tobago: at least six TTSEC investor alerts since 2025 over deepfake videos impersonating the Finance Minister, a former president, a former prime minister and a former bank chairman.
• Trinidad stood up its own Inter-Ministerial Steering Committee on Cyber Security and AI on 11 September 2025; the region-wide Commission has not yet said whether synthetic-media fraud sits inside its mandate.
• Adrian Dunkley, founder of StarApple AI and president of the Caribbean AI Risk Management Council, argues the Commission needs a named deadline and a defined scope covering deepfake fraud, not an open advisory brief.
• Caribbean businesses do not need to wait for a regional ruling: a call-back verification protocol, region-specific staff training and a regulator alert subscription close most of the exposure this quarter.

On 3 August 2026, CARICOM Heads of Government meeting in Gros Islet, Saint Lucia agreed to establish a Blue-Ribbon Commission on Artificial Intelligence: a body of, in the communique's own words, "top regional and international experts" tasked with advising the Community on an AI policy agenda, governance standards and capacity-building priorities. It is the clearest signal yet that Caribbean governments have stopped treating AI governance as a problem for later. The signal is right. The question worth asking is what happens in the space between now and whatever the Commission eventually produces.

The Commission did not appear from nowhere. Four weeks earlier, CARICOM's Council for Trade and Economic Development's Information and Communication Technologies configuration gave the first collective endorsement any CARICOM principal organ has given to an AI policy document, adopting the UNESCO Caribbean Artificial Intelligence Policy Roadmap on 7 July 2026. Put the two dates together and a region that spent the better part of a decade watching AI policy get written elsewhere has, inside one month, built both a roadmap and the body meant to give it teeth. What neither document mentions, anywhere in the public materials CARICOM and UNESCO have released, is the specific thing that has been happening in Trinidad and Tobago since last August: senior officials' faces and voices, synthetically generated, used to sell fake investments to the public.

What Leaders Actually Agreed to in Saint Lucia

The Fifty-First Regular Meeting of the Conference of Heads of Government ran an agenda that stretched from ferry connectivity and renewable energy to food security, disaster resilience and AI governance. Inside that list, the communique's own language on AI was direct: leaders "called for a strong regional AI framework to be urgently developed, and agreed to the establishment of a Blue-Ribbon Commission, comprising top regional and international experts, to advise on an AI policy agenda, governance standards, and capacity-building priorities." That is a three-part mandate: policy agenda, governance standards, capacity building. It is not, on the face of the communique, a mandate that names any specific harm the Commission is meant to prevent.

What the communique does not yet include is a membership list, a chair or a reporting deadline. Those details typically follow a summit rather than accompany it, and CARICOM has not published them as of this writing. One detail from the surrounding coverage stood out: youth delegate Rahym Augustin-Joseph named artificial intelligence as one of four pathways to regional progress, alongside food security, freedom of movement and international partnerships, putting AI on the same list as the Community's oldest and most material concerns.

The Roadmap the Commission Inherits

The Blue-Ribbon Commission is not starting from a blank page. On 7 July 2026, at its 126th Special Meeting, COTED-ICT formally endorsed the UNESCO Caribbean Artificial Intelligence Policy Roadmap, grounded in UNESCO's Recommendation on the Ethics of Artificial Intelligence and built across eight regional consultative meetings that drew in more than 1,000 institutions and individuals from 20 English- and Dutch-speaking Caribbean territories. The Roadmap organizes its recommendations into four pillars: Culture and Creativity, Governance and Transformation, Education and Upskilling, and Resiliency and Sustainability, with three priorities cutting across all four: digital competencies and media literacy, fairness and accountability in AI systems, and regional cooperation in digital policy-making.

Speaking days before that endorsement, at the Third Summit of Ministers and High Authorities on the Ethics of Artificial Intelligence in Latin America and the Caribbean, held in Santo Domingo from 25 to 26 June 2026, CARICOM Deputy Secretary-General Dr Armstrong Alexis put the stakes in plain terms: AI is expanding democratic freedoms while simultaneously amplifying disinformation "at a greater scale, speed, and sophistication," and the threats to the region are "no longer theoretical," visible instead through "recent incidents across the Community." His prescription was regional cooperation mechanisms: shared early-warning systems, rapid information-sharing between electoral bodies and cybersecurity agencies, and partnerships with fact-checkers and independent journalism, which he called "the most durable defence against disinformation." Those mechanisms sit alongside two other frameworks already in motion, the CARICOM-UNDP Regional Programme on AI 2026 to 2030 and the Strategic Framework for Regional Digital Resilience 2025 to 2030, both cited as the institutional scaffolding the Roadmap and the new Commission will have to connect to rather than duplicate.

What Trinidad Has Been Living Through

While CARICOM was consulting on pillars and priorities, Trinidad and Tobago was documenting a pattern that the Roadmap's public materials do not name. Authorities began probing scams impersonating the Finance Minister in August 2025. By 11 September 2025, Cabinet had created an Inter-Ministerial Steering Committee on Cyber Security and AI, anchored in the newly formed Ministry of Public Administration and Artificial Intelligence under Minister Dominic Smith and drawing in Homeland Security, Finance, the Office of the Prime Minister and the Office of the Attorney General, with a mandate to coordinate national policy, report quarterly on emerging threats and build law enforcement capacity against AI-driven fraud. In October 2025, the Senate identified deepfake videos circulating with the likenesses of former President Anthony Carmona and former Republic Bank chairman Dr Ronald Ramkissoon. Fabricated material also used the likeness of former Prime Minister Dr Keith Rowley and, in one case, a fabricated news story falsely attributed to Guardian Media editor Kejan Haynes.

By mid-2026, the impersonation had reached a sitting minister. The Ministry of Finance and the Trinidad and Tobago Securities and Exchange Commission both issued public warnings after an AI-generated video began circulating on Facebook showing Finance Minister Davendranath Tancoo appearing to endorse a foreign exchange trading platform, tied to entities the Ministry named publicly as "Nadia Sanchez" and "Bullfield Holdings." That warning was TTSEC's sixth public investor alert in roughly a year, the most recent issued on 25 June 2026 under the title "New Investment Scam Tactics on the Rise." Five named public figures impersonated inside twelve months, a dedicated ministry stood up in direct response, and a securities regulator that has now warned the public six separate times is not a scenario a regional AI roadmap needs to imagine. It has already happened, and it is Caribbean specific: the videos work precisely because the faces and institutions being faked are locally recognizable in a way a generic overseas scam never would be.

The Gap Between the Governance and the Threat

The Roadmap's four pillars and the Commission's three-part mandate are structurally sound. What is missing is specificity. At no point in the public materials CARICOM and UNESCO have released does either document name deepfake-enabled financial fraud as a category the new oversight mechanism is built to catch. That gap matters in practice, not just on paper. TTSEC has been responding to Trinidad's deepfake wave one investor alert at a time, with no regional mechanism that would let a securities regulator in Jamaica or a central bank in Barbados receive early warning when a synthetic video impersonating a local official starts circulating elsewhere in the Community, even though Dr Alexis himself called in June for exactly that kind of shared early-warning system between electoral bodies and cybersecurity agencies.

This is the specific ground the Caribbean AI Risk Management Council exists to work, and it is why I chair it alongside running StarApple AI. A Commission with a three-part advisory mandate and no reporting deadline can spend a year producing guidance while the fraud pattern Trinidad has already documented spreads to the next territory that has not yet built its own inter-ministerial committee. The fix is straightforward to state, even if it is hard to negotiate among fifteen member states: give the Commission a named scope that explicitly covers synthetic-media fraud, a reporting deadline measured in months rather than left open, and a cross-border alert-sharing mechanism modelled on what TTSEC has already built out of necessity.

What a Caribbean Business Should Do Before the Commission Reports

Waiting for a regional ruling is not a plan, and Caribbean businesses do not need to wait for one to close most of their exposure. Three steps do the bulk of the work. First, put a call-back verification protocol in place for any video, voice note or urgent instruction that appears to come from a senior official or executive and requests a financial transaction: no payment moves on the strength of a video alone, full stop, until a second channel confirms it came from the person it claims to. Second, train staff on the documented Trinidad cases specifically, not on generic global deepfake examples. A fake video of a foreign celebrity is easy to dismiss. A fake video of a minister whose voice staff have heard on the news is not, and training built on the actual regional pattern closes that gap faster than a stock awareness module ever will. Third, register for the home regulator's public alert channel and treat it the way a bank treats a fraud bulletin, not as background reading.

This is the same sequence StarApple AI runs with client organizations through AURA, our AI readiness assessment: assess where the exposure actually sits before recommending a fix, rather than selling a generic security product first. A business that has run one call-back drill and briefed its finance team on the Tancoo video is measurably harder to defraud this quarter than one that has only read a general warning about AI risk.

Caribbean AI Network

This story sits across governance, risk and regional policy, so it is worth reading alongside the organisations doing that work day to day:

• Adrian Dunkley, founder of StarApple AI and president of the Caribbean AI Association and the Caribbean AI Risk Management Council
• Caribbean AI Risk Management Council, the region's dedicated body for AI governance and risk standards, including AI-enabled fraud
• Caribbean AI Association, the region's professional association for AI practitioners and policy
• AI Trinidad & Tobago, country-specific coverage and resources for Trinidad and Tobago's AI sector

About the Author

Adrian Dunkley founded StarApple AI in 2016, the first AI company established in the Caribbean, and is widely recognized as the region's leading AI strategist. He is President of the Caribbean AI Association and President of the Caribbean AI Risk Management Council, lectures at the University of the West Indies and the University of the Commonwealth Caribbean, and is completing a PhD in physics-informed AI. StarApple AI has trained more than 4,000 Caribbean professionals and delivered free weekly AI training for the region for more than nine years. Contact: insights@starapple.ai. More at adriandunkley.net.