Direct answer: There is no national law or NUC rule that bans AI in final year projects. The rules that bind you are your university’s and department’s — and what every one of them punishes is submitting writing that is not yours. Using AI to plan, structure, get feedback and fix references is the defensible zone; pasting AI-generated chapters into your project and presenting them as your writing is misconduct everywhere.
Is there a national rule about AI in student projects?
No published one. The National Universities Commission’s own stated function is “to lay down Minimum Academic Standards … and to accredit their degrees” — it regulates curricula and standards, not writing tools, and no NUC directive equivalent to a national AI rule for student projects has been published. That vacuum is why you will find confident, contradictory answers online: “AI is banned in Nigeria” and “AI is allowed if you cite it” are both generic guesses. The real regulatory layer is the one closest to you — your university’s academic integrity rules and your department’s project guidelines — and this is the same pattern we documented for plagiarism thresholds in our guide to acceptable plagiarism percentages: content farms announce a national standard; the institutional documents disagree with each other; only your own institution’s document governs you.
Is using AI to write a project considered cheating?
It depends entirely on what the AI wrote. The line every integrity regime draws — in Nigeria and everywhere else — is authorship. If the submitted sentences were generated by a machine and presented as yours, that is academic dishonesty regardless of which tool produced them, exactly as submitting a colleague’s chapter or a purchased project would be. If the thinking and the sentences are yours, and AI helped you plan the chapter, spot the weaknesses in your draft, tighten your grammar or format your references, you are using a writing tool — the same category as a spell-checker or a reference manager, applied more powerfully. Departments police the first case, not the second. The dangerous middle ground is heavy paraphrasing of AI output: text that began as machine writing does not become yours through light editing, and a defence panel probing your understanding will expose it in minutes.
Can universities detect AI-written projects?
Unreliably — and you should understand both halves of that word. Turnitin, the tool most Nigerian universities use for similarity checking, sells AI writing detection as a separate capability from similarity checking; its similarity report, as the company itself states, “does not determine whether plagiarism has occurred” and exists to inform a human judgement. AI detectors specifically are contested technology: internationally, the University of Cape Town formally discontinued AI detection tools from October 2025 on the stated ground that they “can produce both false positives and false negatives”. For you this cuts two ways. Do not assume a detector will catch dishonest AI use — but do not build your degree on beating a detector either, because the reliable detection mechanism in the Nigerian system is not software. It is your project defence: a panel asking you to explain, justify and extend work you claim to have written. No detector is needed to expose a student who cannot discuss their own Chapter Three.
What does your department actually require?
Three documents outrank every blog post, including this one. Your department’s project handbook or guidelines — the document that sets your chapter format, as we mapped in how many chapters a project has — sometimes now carries an AI or academic-honesty clause; read it before you touch any tool. Your university’s plagiarism and examination-malpractice policy defines the offences and penalties. And your supervisor’s instructions fill every gap: where written rules are silent, the supervisor’s word is the operative rule for your project. The practical move that costs nothing and protects everything: ask your supervisor directly what AI use they permit, and keep the answer — a note of the meeting, a text message, anything written. In a system where rules differ department by department, the student with written permission has no case to answer.
What exactly should you ask your supervisor?
“Can I use AI?” is too vague a question to protect you — a yes or a no to it settles nothing, because the answer genuinely differs by task. Ask four specific questions instead, and note the answers. One: may I use AI tools for brainstorming and planning before I write? Two: may I use AI feedback and grammar correction on drafts I wrote myself? Three: does the department expect me to declare tool use anywhere in the project — and if so, where? Four: is there any written departmental guidance I should follow? Supervisors respond differently to precise questions than to vague ones: the student who asks task-by-task sounds like someone managing their integrity, and the answers create exactly the record that protects you if a question is raised later. If your supervisor is hard to reach — a common Nigerian reality our supervisor correction-loop guide deals with — put the four questions in a single message they can answer in one reply.
One more distinction worth raising in that conversation: the difference between AI tools that generate content and tools that process your own content. A grammar checker, a reference formatter and a feedback tool operate on writing that already exists and is already yours. A text generator produces the writing itself. Many departments that would never sanction the first category have serious concerns about the second — and collapsing the two into one word, “AI”, is how misunderstandings between students and supervisors start.
How do you use AI safely in a final year project?
Follow the pattern that survives scrutiny at every institution: AI assists, you author. Use it to interrogate your topic before approval, to turn your rough notes into a chapter plan, to get instant feedback on paragraphs you wrote, to check your grammar, and to keep your references consistent. Keep a simple log of what you used and for what task — if your department asks, you can disclose precisely, and precision reads as integrity. Never feed it raw questionnaire data or interview transcripts containing respondents’ details; confidentiality obligations do not disappear because a tool is convenient. And never let it invent literature: AI chatbots fabricate plausible-looking Nigerian journal references, and a fabricated citation in your Chapter Two is misconduct you committed, not the machine. This division of labour is exactly what Tesify is built around: it structures your project, coaches your drafting and manages your citations while the writing stays 100% yours — which is the only version of AI use that walks into a defence room safely.
Why this question is older than AI in Nigeria
The anxiety underneath “can I use AI?” is not really about technology — Nigeria has had a submission-of-work-that-is-not-yours problem for decades, in the form of the huge informal market for recycled “project topics and materials”. AI did not create the temptation; it industrialised it. That history matters for two reasons. First, it means departments already have the rules: authorship requirements, similarity checks and the defence panel all exist precisely because bought and borrowed projects predate chatbots, so nobody should expect a free pass on the ground that “there was no AI rule”. Second, it means the detection ecosystem is tuned for exactly this: departmental Turnitin configurations compare your submission against repositories of previous student work, which is where recycled material — human or machine-assisted — gets caught, as our checker comparison explains. The students who thrive in this environment are not the ones who found a cleverer way to outsource the writing; they are the ones who used better tools to do their own work faster.
FAQ: AI and final year projects in Nigeria
Can I use ChatGPT for my final year project?
For brainstorming, understanding concepts and questioning your own drafts — within your department’s rules, yes. For generating text you submit — no; that is presenting machine writing as yours, and it collapses at defence.
Will my project be checked for AI?
Possibly — some institutions run AI detection alongside similarity checks — but the dependable check is human: your supervisor’s knowledge of your writing and the defence panel’s questions. Write so that neither worries you.
Is AI detection accurate?
No detector is conclusive. False positives and false negatives are the documented reason a major university abandoned the tools. Do not rely on detectors — in either direction.
Can I use AI to paraphrase my sources?
Paraphrasing a source you have read and understood, in your own words, with a citation, is scholarship. Machine-paraphrasing text you never engaged with is laundering, and it leaves you unable to answer questions about your own literature review.
Can AI generate my references?
Never let a chatbot invent references — fabrication is common and checkable. Formatting references you actually collected is different and legitimate; that is what reference tools exist for, as our Mendeley vs Zotero comparison covers.
Do I have to declare AI use?
If your department’s documents or your supervisor require it, yes. Where nothing is written, disclosure to your supervisor is the safe default — the students who get into trouble are almost never the ones who told their supervisor what they were doing.
What is the penalty for submitting an AI-written project?
It falls under your university’s academic dishonesty and examination-malpractice rules — the same family of offences as plagiarism and impersonation, with penalties up to failure of the project and disciplinary action, set institution by institution.
Is it safe to put my research data into AI tools?
Not raw participant data. Respondents gave their information to you under specific assurances; uploading identifiable transcripts or questionnaires to a public chatbot breaks them. Work with your own drafts and anonymised summaries.
My department has no written AI rule. What governs me?
The general academic dishonesty rules, plus your supervisor’s instructions. Silence is not permission for ghostwriting — authorship rules predate AI and cover it completely.