Beyond the AI Panic: Rethinking Higher Education in Sierra Leone for an Age of Artificial Intelligence
Artificial intelligence is no longer approaching the university. It is already inside it.
Students use it to research, write, translate and solve problems. Lecturers use it to prepare teaching materials, analyse information and support academic writing. Administrators use it to process data, while employers increasingly reorganise work around it.
For Sierra Leonean universities, therefore, the question is no longer whether AI should enter higher education. That debate has been overtaken by reality. The more important question is: How should our universities embrace AI without compromising educational quality, academic integrity, human judgement and students’ capacity to think for themselves?
Recent commentaries in University World News illuminate different dimensions of this challenge. Antoine Casanova-Mazet examines AI, curriculum and labour-market demands. Phuong Nguyen focuses on forms of reasoning universities must preserve. James Yoonil Auh interrogates academic labour and intellectual property. Yan Li and George Wood raise questions of fairness and transparency in AI-assisted admissions, while Fredrick Otike warns against African universities becoming dependent consumers of knowledge produced elsewhere.
Together, their arguments suggest that AI is not merely another educational technology. It is reshaping how knowledge is produced, how learning is demonstrated and what it means to be educated.
Beyond the cheating debate
The immediate university response to generative AI has understandably centred on cheating. Did the student write the essay? Was ChatGPT used? How can lecturers detect AI-generated work?
These concerns are legitimate, but too narrow. AI can already produce competent essays, reports and presentations. Assessment systems based principally on polished final products will therefore become increasingly difficult to defend.
Casanova-Mazet argues that assessment should shift towards process and judgement. Educators should examine how students frame problems, select evidence, challenge AI-generated information and justify their conclusions.
This matters enormously for Sierra Leone. A student who submits an impressive AI-generated essay may demonstrate little intellectual development. Yet one who uses AI, identifies its weaknesses, verifies its claims and develops an independent argument may demonstrate precisely the higher-order thinking universities should cultivate.
Academic integrity in the AI era must therefore move beyond asking, “Did the student use AI?” We must increasingly ask: “What intellectual work did the student personally perform?”
Protect the thinking that must remain human
AI is increasingly capable of identifying patterns, summarising knowledge and applying established frameworks. Yet universities often reward students for precisely these activities.
Writing in University World News, Phuong Nguyen argues that graduate education should cultivate abductive reasoning: the capacity to encounter something unexpected, recognise that established explanations are inadequate and generate a new explanation.
For Sierra Leone, this is critical. AI can provide information and suggest solutions to youth unemployment, public-sector reform, climate vulnerability, revenue mobilisation, agricultural productivity and educational quality. But Sierra Leone still needs graduates capable of asking: Will that solution work here?
That requires context, experience and judgement.
Nguyen proposes that learners establish their own position before consulting AI and identify what evidence could change it. Otherwise, the machine’s first coherent interpretation can easily become the student’s own.
Our educational principle should therefore be simple: think first, consult AI second, interrogate its response, and defend your own conclusion.
Curriculum reform cannot stop at teaching ChatGPT
Universities must resist the temptation to add an “AI course” to existing programmes and consider the problem solved.
Casanova-Mazet observes that AI is changing tasks and capabilities within occupations, even where job titles remain unchanged. Universities must therefore identify which capabilities are becoming more valuable, which are being automated and which remain fundamentally human.
His concept of “curriculum intelligence” is particularly relevant. AI-supported analysis could help universities examine job advertisements, employer feedback, internship evaluations, graduate outcomes and emerging skills against existing curricula.
But algorithms must not determine what universities teach. Labour-market intelligence should inform academic judgement, not replace it. Faculty, employers, professional bodies, students and society must remain central to curriculum decisions. A university, after all, is not merely a labour-market training centre.
Integrity must apply to institutions too
Universities cannot demand transparency from students while remaining opaque about their own use of AI.
Yan Li and George Wood identify an emerging accountability asymmetry in university admissions: applicants may face detailed restrictions on AI use while institutions disclose considerably less about their own deployment of AI. They emphasise fairness, accountability, transparency, ethics and meaningful human oversight.
The lesson extends beyond admissions. If Sierra Leonean universities use AI in grading, plagiarism investigations, student progression, staff appraisal or other high-stakes decisions, those affected should know, and such decisions should remain subject to meaningful human review.
Academic integrity cannot be demanded only from below. Institutional integrity must accompany student integrity.
AI should augment, not diminish, the academic
James Yoonil Auh argues that AI need not replace professors to transform their profession. It can redistribute tasks involving teaching preparation, feedback, advising, research and administration.
Used wisely, this could be beneficial. Time saved on repetitive tasks could return to research, mentoring, supervision and meaningful interaction with students.
But AI also creates what Auh describes as “AI verification labour.” Generated material must be checked. Citations must be verified. Bias and fabricated information must be detected. Assessments must be redesigned.
Universities should therefore avoid turning AI-driven efficiency into larger workloads and less human interaction. Technology should remove unnecessary bureaucracy, not automate the intellectual relationships at the heart of education.
Intellectual property is equally important. Recorded lectures, slides, assessments and academic feedback can become training material for AI systems. Sierra Leonean universities therefore need clear rules on ownership, consent and compensation before technological capability overtakes institutional policy.
Africa must not become merely an AI consumer
Perhaps the most consequential warning comes from Fredrick Otike’s description of a “digital colonial time bomb.”
Writing about Kenya, Otike warns that African universities risk becoming consumers of knowledge controlled and monetised elsewhere. AI deepens that danger. If African scholars do not publish research, write books, digitise indigenous knowledge and document local experiences, African perspectives will remain marginal in the emerging AI knowledge economy.
The warning applies powerfully to Sierra Leone.
What happens when AI knows more about European public administration than Sierra Leonean local governance? What if it understands Western labour markets but little about our informal economy, chiefdom governance, indigenous knowledge and historical experience?
Our universities must therefore become producers, not merely consumers, of digital knowledge.
A 10-Point Framework for Responsible AI in Sierra Leonean Universities
The response must be practical rather than rhetorical.
- Establish institutional AI governance. Create multidisciplinary structures involving faculty, administrators, ICT professionals, librarians and students.
- Adopt clear AI-use policies. Define permitted, restricted and prohibited uses, with appropriate disclosure requirements.
- Redesign assessment. Emphasise reasoning, oral defence, practical projects, reflective work, supervised assessment and locally grounded problem-solving.
- Make AI literacy a graduate capability. Teach verification, source evaluation, bias detection, privacy, ethics and responsible use—not merely prompting.
- Protect human judgement. Admissions, grading, disciplinary action, progression and other high-stakes decisions should never be surrendered entirely to algorithms.
- Build faculty capacity. Provide continuing professional development in AI-supported pedagogy, assessment, research integrity and verification.
- Protect academic intellectual property. Establish consent and ownership rules before faculty-generated materials, voices, images or research are used to train AI systems.
- Create continuous curriculum intelligence. Monitor changing labour-market capabilities while retaining academic control over curriculum decisions.
- Build Sierra Leone’s digital knowledge base. Strengthen institutional repositories, local journals, books, datasets, theses and responsibly documented indigenous knowledge.
- Measure AI by educational value. Success should be judged not by the number of AI tools acquired, but by improvements in learning, research, access, academic integrity and meaningful human engagement.
The real test
The AI revolution in Sierra Leonean higher education will not be measured by how many universities acquire AI platforms or how many students master prompting.
The deeper measure is whether our universities produce better thinkers because AI exists.
We must move beyond asking whether students use AI and ask whether they are learning with it without surrendering their capacity to think without it.
AI can make information abundant, intellectual production faster and convincing answers cheap. But universities have never existed merely to manufacture answers.
Their higher purpose is to develop people capable of deciding which questions matter, which evidence deserves trust, which answers are inadequate, and what ought to be done next.
That is where academic integrity now begins.
And that is where the Sierra Leonean university must remain indispensable.
Chernor Mohamadu Jalloh is a Lecturer in Governance and Development Studies at the Institute of Public Administration and Management (IPAM), University of Sierra Leone. A Fulbright Scholar and policy analyst, his research and writing focus on governance, public policy, development, higher education, institutional reform and sustainable development.
This article is a commentary. The views expressed are those of the author and do not necessarily represent those of IPAM or the University of Sierra Leone.