92% of Students Are Using AI: The Literacy Gap Caribbean Youth Can Close First
A new global survey says almost every student already uses AI. Almost none of them were properly taught how. That gap is an opening, and Caribbean youth are in a position to close it before the rest of the world does.
A global survey published this year by the Digital Education Council put a number on something teachers across the Caribbean have been saying for two years without proof: 92% of students and 79% of faculty are now actively using AI in their coursework. The study drew on more than 45,000 responses from 27,284 students and 18,114 faculty members, spread across 35 countries. Two years ago a number like that would have made a headline on its own. Now it just describes how a normal classroom works.
Buried lower in the same report is the number that actually matters: only 29% of students believe their instructors are equipped to guide them on AI use. Adoption ran way ahead of instruction, and nobody has closed the distance between the two.
The AI literacy gap is simple: nearly all students already use AI, but formal teaching on how to use it well lags far behind, especially for younger learners. For Caribbean youth, closing that gap early turns AI from a study crutch into a career skill the region can export to the world.
The News: A Global Survey Just Confirmed What Teachers Already Suspected
The Digital Education Council's AI in Higher Education Global Survey 2026 is one of the largest datasets ever collected on how students and faculty actually use AI, not how policy documents say they should. The topline figures are striking on their own. Ninety two percent of students and seventy nine percent of faculty are actively engaging with AI. But look at how that engagement breaks down and the picture gets messier. Only 15% of students say AI shows up in many of their courses in any structured way. Forty three percent say it appears in a few courses. Another 43% say it has not been integrated at all, meaning they are using it on their own, unsupervised, unguided, figuring it out by trial and error.
A separate finding, drawn from Education Week reporting and cited in a 2026 review of education data, sharpens the picture further by age. Only 8% of students in grades Pre-K through 3rd receive formal AI literacy lessons. Eighty percent of high school students, according to their own teachers, do. The youngest learners, the ones forming their very first habits around a tool that is not going anywhere, are the least likely to be taught anything about it at all.
None of this means students are lazy or reckless. A UK-focused Student Generative AI Survey from HEPI, based on 1,054 full-time undergraduates surveyed by Savanta in December 2025, found 95% of students using AI in at least one way, and 49% saying it genuinely improved their student experience through time saved and clearer understanding. The schools around them, secondary schools especially, are the ones who have not caught up to what students were already doing on their own.
🧠Why "AI literacy" means something different at 8 than it does at 17
The 8% versus 80% gap is not just a resourcing failure, it lines up with how thinking develops. A seven or eight year old is still firmly in what Piaget called the concrete-operational stage: she reasons brilliantly about things she can see, touch, and test, but abstract rules about "when is it fair to use a tool that thinks for you" do not land yet. What she can absorb at that age is concrete and habitual: AI can get things wrong, always check with a grown-up, a chatbot is not a friend. Small, repeatable rules, not a lecture on ethics.
By the time a student reaches secondary school, formal-operational thinking has usually come online. A CSEC or CAPE student can reason about hypotheticals, weigh trade-offs, and hold two competing ideas in their head at once, which is exactly the kind of thinking that a real AI literacy conversation requires: when does AI help me think better, and when does it think for me instead? That is precisely why the 80% figure for high schoolers makes sense and the 8% figure for young children does not. The lesson has to change shape as the learner does, not just get simpler or louder.
Piaget's stagesConcrete vs. abstract reasoningHabit formationScaffolded literacy
Why This Hits Different in the Caribbean
A student in Toronto or London who falls into that 43% with no AI integration in class still has a wide safety net: better broadband on average, more teachers who received AI training as part of continuing education, more free after-school programmes competing for their attention. A student in rural St. Elizabeth or the hills above Gros Islet does not automatically have that same net. If the only AI education a Caribbean teenager gets is what they figure out alone at 11pm trying to finish a School-Based Assessment, they are learning the tool in the least structured, least critical way possible, exactly the version of AI use that gets flagged, that produces shallow work, and that leaves them unable to explain their own thinking when a teacher asks them to.
Wi cyaan mek dat di norm. The Caribbean Examinations Council has already moved on this. CXC's 2026 policy on AI in School-Based Assessments introduced a Disclosure Form and an Originality Report requirement, which forces every student to say exactly how and where they used AI in their own work. That policy only works if students were taught, ahead of time, how to use AI in a way they can honestly disclose. A student who never learned the difference between using AI to check their work and using AI to produce their work is set up to fail that disclosure, not because they cheated on purpose, but because nobody showed them where the line was.
By the numbers
Digital Education Council, AI in Higher Education Global Survey 2026 (35 countries, 45,398 responses)
Trinidad Just Put a Regional Flag in the Ground
The good news is that the region is not waiting to be handed a plan from somewhere else. In July 2026, the first Caribbean Artificial Intelligence Forum convened in Trinidad and Tobago, bringing together the Caribbean Telecommunications Union, The University of the West Indies St Augustine, the Artificial Intelligence Innovation Centre, regional policymakers, academics, and youth representatives. The forum's Deputy Secretary-General, Nigel Cassimire, put the stakes plainly: AI has to be harnessed to reduce the region's vulnerabilities and amplify its strengths, which requires standards that fit Caribbean scale and governance that protects Caribbean citizens. Dr Craig Ramlal added the sharper version: the Caribbean cannot afford to sit out the AI revolution, because together the region has the scale to build harmonised policy that no single island could manage alone.
The forum produced a Caribbean AI Task Force report with policy recommendations that go beyond boardrooms and government offices, touching education and workforce readiness directly. That matters because it means the conversation about AI literacy is no longer confined to individual schools improvising on their own. It is becoming a regional priority, with youth representatives sitting in the room where those decisions get shaped.
"Mi see plenty a mi students use AI like a vending machine, put in a question, tek out an answer, nuh think about it again. Di ones who actually get better, dem use it like a sparring partner. Dem argue wid it, check it, push back. Dat difference is di whole ballgame, and it's teachable." - Kimoy Anderson, AI Literacy Mentor, The Genius Project
Five AI-Literacy Habits Every Caribbean Teen Should Build This Term
Closing this gap does not require a new subject on the timetable or a school budget nobody has. It requires a handful of habits, taught early and repeated often.
- Treat AI as a sparring partner, not a vending machine. Ask it to explain its reasoning before you accept an answer. If it cannot justify a step, do not use that step.
- Check everything against a second source. AI tools produce confident, fluent answers that are sometimes wrong. The habit of verifying, especially for dates, statistics, and CXC syllabus content, is the single most protective skill a student can build.
- Know your disclosure rules before you start, not after you finish. Under CXC's SBA policy, you need to be able to say exactly how AI was involved in your work. Build the habit of tracking that as you go.
- Use AI to practise, not to produce. Feeding it past-paper questions and asking it to quiz you builds understanding. Asking it to write your final answer builds nothing you can defend in an exam room.
- Build one real project, not ten shallow prompts. A single AI-assisted project, a chatbot, a data model, a tool that solves something in your own community, teaches more than a hundred one-off homework shortcuts ever will.
How The Genius Project Fits In
The Genius Project exists specifically to close this gap, tuition free, for Caribbean youth. Founded by Adrian Dunkley, the programme gives students hands-on AI education and mentorship instead of leaving them to figure the tool out alone at 11pm before a deadline. Dunkley built StarApple AI, the first AI company to launch anywhere in the Caribbean, and he is widely regarded across the region as its leading voice on what AI can do and what it takes to teach it responsibly. He started The Genius Project so that expertise would reach a fifteen year old in Half Way Tree or a sixth former in Bridgetown just as directly as it reaches a government briefing.
The programme does not sit alone. It is part of a wider Caribbean AI ecosystem that includes the Caribbean AI Association, which works on the same regional policy questions raised at the Trinidad forum, and StarApple AI, which provides the professional-grade expertise that shapes what The Genius Project teaches. For more on Adrian Dunkley's wider work across the region, see adriandunkley.net.
What Comes Next
The gap this article opened with, 92% usage against 29% trust in guidance, closes only if individual schools, mentors, and programmes actually teach the habits above instead of assuming students will pick them up by osmosis. The Caribbean AI Task Force report gives the region a shared starting point. CXC's disclosure requirement gives students a concrete reason to build the habit now rather than later. And free programmes with real mentorship give every Caribbean teenager, not just the ones whose schools already have the budget, a real way in.
The students who learn to argue with AI instead of just asking it questions are the ones who will walk into university, into CXC results day, and into their first job interview already ahead. You can already find that happening in a handful of classrooms and mentoring sessions around the region, well ahead of whatever policy eventually catches up to it.
See also: AI, CXC, and the SBA: A Caribbean Student's Guide, Summer of AI: Why 2026 Is the Year Caribbean Youth Become the Region's Most Powerful Tech Asset, and 5 Myths About AI That Hold Young People Back.