The Caribbean Examinations Council (CXC) should be commended for taking decisive steps to safeguard the integrity of its assessments in response to the rapid advancement of artificial intelligence (AI). Maintaining public confidence in Caribbean qualifications is essential, and CXC’s willingness to make difficult decisions demonstrates its commitment to ensuring that the qualifications it issues continue to command the confidence of all stakeholders.

There is no doubt that AI has fundamentally changed the educational assessment landscape. With tools now capable of generating essays, analysing texts, solving complex problems, and producing sophisticated responses in seconds, concerns regarding the authenticity of school-based assessments (SBAs) are both understandable and justified. CXC’s decision to review its assessment model reflects responsible leadership in a time when educational systems worldwide are grappling with similar challenges.

However, while protecting assessment integrity is essential, the irresponsible AI use by learners should not signal a retreat from SBA assessments; rather, it should serve as a catalyst for redesigning the assessments in ways that continue to develop and evaluate the knowledge, skills, and competencies that matter most in this ever-changing world.

The SBA component has enabled students to investigate authentic problems, apply knowledge to real-world situations, conduct basic research, communicate ideas, and engage in critical thinking beyond what is typically possible in a traditional written examination. These are precisely the competencies that education systems across the world continue to identify as essential for success in the 21st century.

Rather than replacing the SBA for many subjects with a paper three, an assessment still completed under examination conditions, perhaps this is the ideal opportunity for CXC to lead the Caribbean in developing new assessments that are both AI-resilient and educationally meaningful.

A paper three cannot fully replicate the richness of authentic assessment experiences that require sustained inquiry, creativity, application, and reflection. Therefore, the question before us is not simply how to AI-proof assessment, it is how to design assessments that prepare students for the world they are living in today and the one they will help shape tomorrow. Our assessment systems must evolve alongside the realities of work, society, and technology, while continuing to develop the human qualities that remain indispensable.

Authentic assessment should move beyond simply judging the final product that students submit. Instead, it should place greater emphasis on the learning journey and the thinking behind the work. Every SBA could conclude with a brief oral presentation or structured conversation in which students explain their thinking, justify their decisions, respond to questions, and reflect on their learning.

This simple addition would require students to demonstrate genuine ownership of their work, making their thinking visible in ways that AI cannot replicate and teachers can confidently verify. It would also help develop the communication, critical thinking, and presentation skills that students are increasingly expected to demonstrate in higher education and the modern workplace.

Students could also be required to document their learning through research journals, design logs, planning documents, reflective commentaries, annotated drafts, digital portfolios, and structured teacher conferences. These artefacts would provide evidence not only of what students know but also of how they arrived at that knowledge. When assessment values reasoning, reflection, revision, and application alongside the final product, opportunities for inappropriate AI use are substantially reduced.

Artificial intelligence challenges us to reconsider what we value in assessment.

Importantly, this approach also aligns with the principles of inclusive assessment. As assessment scholars remind us, fairness is not achieved by treating every learner identically, but by designing assessments that provide equitable opportunities for all students to demonstrate the intended learning outcomes. Inclusive assessment begins at the design stage by anticipating learner diversity rather than relying solely on accommodations after barriers emerge.

Authentic assessment supports this philosophy by allowing students to demonstrate their learning through multiple forms of evidence while maintaining common academic standards and expectations. It also recognises an important reality that has not changed simply because AI has emerged.

Exam anxiety remains a genuine barrier for many capable students. While examinations will always have an important place in education, they should not be the sole or dominant measure of achievement. By combining written work with oral presentations, reflective discussions, and opportunities for students to explain and defend their thinking, we can create assessments that are not only more resistant to inappropriate AI use, but also more inclusive, more representative of learners’ abilities, and better aligned with the diverse ways in which learners demonstrate understanding.

AI also challenges us to reconsider what we value in assessment. If AI can produce a competent essay within seconds, then written responses alone can no longer be regarded as sufficient evidence of learning. Instead, our assessments should intentionally cultivate and evaluate the competencies that remain uniquely human: critical thinking, creativity, communication, collaboration, citizenship, and character — the six Cs that underpin deep learning.

These competencies cannot be authentically outsourced to technology. They require learners to question, analyse, create, defend ideas, solve problems, work with others, and exercise ethical judgment in authentic contexts. Our responsibility as educators extends beyond safeguarding academic integrity. It includes explicitly modelling and teaching students how to use AI ethically, critically, and responsibly as a tool to support learning, inquiry, reflection, and creativity — not as a substitute for thinking.

Just as calculators eventually found their appropriate place within mathematics education, AI must similarly find its place within teaching, learning, and assessment. Students should be encouraged to acknowledge and critically evaluate their use of AI rather than conceal it.

CXC has rightly reminded us that the integrity of its qualifications is non-negotiable. I wholeheartedly agree. However, preserving integrity should not come at the expense of authentic learning. The future of assessment should not be determined by what is easiest to secure under examination conditions but by what most effectively prepares learners for an increasingly complex, collaborative, and technology-rich world.

The Caribbean has an opportunity to become a global leader in assessment innovation. Let’s rise to the challenge.

Shellon Samuels-White is a student assessment officer at The Mico University College. Send comments to the Jamaica Observer or shellon.samuels-white@themico.edu.jm.