Cheating on Caribbean regional exams jumped sharply this year, and for the first time, some of it involved AI. During the official release of the May-June 2026 CXC results in Anguilla, Director of Operations Dr. Nicole Manning reported 128 confirmed examination irregularities across CSEC and CAPE candidates, up from 80 in 2025, 54 in 2024, and 36 in 2023. Of this year’s total, 106 came from CSEC candidates (more than double 2025’s 57) and 22 from CAPE candidates. The most common violations were unauthorized devices, candidate collusion, and unauthorized materials, and CXC confirmed this was the first year it recorded official cases of AI misuse specifically. Students caught in any of these breaches face grade cancellation or disqualification.
Is the Curriculum the Real Problem?
CXC’s response leans heavily on catching students: more device restrictions, more scrutiny during sittings, harsher consequences. What it doesn’t address is why a doubling of CSEC irregularities happened in the same year AI tools became widely accessible on every student’s phone. CSEC and CAPE exams are still built almost entirely around timed, closed-book recall of memorized material. That format was designed for an era when the only alternative to knowing an answer was leaving the space blank. It was never built to survive a world where a free chatbot can generate a plausible-sounding answer to almost any prompt in seconds.
None of that excuses cheating, and it shouldn’t. Getting caught still means losing your grade, and for CAPE candidates in particular, that can mean losing a university placement. But treating this purely as a discipline problem, more monitoring, harsher penalties, ignores that the incentive structure pushing students toward shortcuts hasn’t changed even as the tools available to exploit it have gotten dramatically more capable.
What CXC Is Actually Doing Differently
To be fair, CXC isn’t standing entirely still. This year’s sitting also introduced a new modular certification track and expanded electronic assessment options, including a hybrid e-assessment pilot where students viewed questions on-screen but still wrote answers by hand. The council also extended compassionate consideration to 359 candidates facing circumstances beyond their control, which suggests at least some recognition that a purely punitive approach isn’t the whole answer.
Why This Matters Beyond One Exam Cycle
A 60 percent year-over-year jump in irregularities isn’t a rounding error, and it’s not going away on its own. AI tools are only going to get more capable and more embedded in how students do everyday coursework, not less. If the assessment format doesn’t evolve alongside that reality, whether through more oral components, project-based assessment, or exams that test reasoning under supervision rather than pure recall, the irregularity count is very likely to keep climbing, along with the number of students whose futures get derailed by a single bad decision made under pressure.
What do you think? Should regional exam boards redesign testing formats around AI’s existence, or is stricter enforcement the right response? Let us know your thoughts in the comments on BeezLoop.com!
