Education7 min read

How Do University Lecturers Use AI?

On this page

University lecturers use AI to draft course outlines, build marking rubrics, write varied practice and exam questions, prepare feedback comments, and handle research admin such as grant forms and meeting notes. The lecturer stays the expert: every draft is checked for accuracy, the marks stay with the marker, and student work and records stay out of public AI tools unless the university has approved a tool for them.

This guide is for lecturers, tutorial fellows and heads of department in Kenyan universities and colleges. It gives seven tasks with prompts you can copy, what to check, and the integrity rules that matter. If you supervise postgraduates, read our guide on using AI in postgraduate research without crossing the integrity line.

Start with your university's AI policy

Your university's own AI policy, and the rules for each course, decide what you and your students may do. If yours has none yet, ask your department or quality assurance office what applies, and tell students in writing what is allowed on each assignment.

UNESCO's Guidance for Generative AI in Education and Research notes that new versions of public AI tools are coming out faster than national rules can adapt. It says the lack of national rules in most countries leaves users' data privacy unprotected and educational institutions "largely unprepared to validate the tools". It calls for a human-centred approach, including protecting data privacy.

Data protection is a law, not just good practice. Under Kenya's Data Protection Act, 2019, student records and marks are personal data, and health details count as sensitive personal data. Section 25 says personal data should be "limited to what is necessary" for its purpose. Section 35 gives people the right, with some exceptions, not to be subject to a decision based solely on automated processing that significantly affects them. In our view, a grade set by an AI tool with no real human judgement could be that kind of decision.

Seven tasks where AI helps, with prompts

Each prompt follows the four Cs we teach at FuKazee: Clarity, Context, Constraints and Customisation. Replace the square brackets. Never paste student names, registration numbers or real scripts into a public tool.

1. Draft a course outline

I teach [course name and code], a [number]-credit [undergraduate / postgraduate] course in [department] at a university in Kenya. Draft a 14-week course outline with: course purpose, five learning outcomes using clear action verbs, weekly topics, one learning activity per week, assessment plan with weights, and a short statement on permitted AI use. Constraints: use only the topics in the approved course description below. Do not invent readings. Where a reading is needed, write [READING]. DESCRIPTION: [paste]

Check: the outline matches the approved course description and your university's format. Add real readings yourself. AI tools often invent references that look real.

2. Build a marking rubric

Create an analytic rubric for this assignment: [paste brief]. Criteria: [list, for example argument, use of evidence, structure, referencing]. Four levels from "excellent" to "below expectations", with a clear description at each level that a second marker could apply the same way. Weight the criteria as [weights]. Plain English that students can understand.

Check: apply the rubric to two scripts you have already marked. If it gives different results from your judgement, adjust the descriptors.

3. Write varied questions at different levels

Write 15 questions on [topic] for [year] students in [course]. Five that test recall, five that test application to a Kenyan case, and five that test analysis or evaluation. For each, give a model answer in under 80 words and the learning outcome it tests. Avoid questions that can be answered by copying a definition.

Check: confirm every model answer is correct. AI makes confident mistakes in specialist subjects. For exams, follow your university's moderation and security rules, and keep draft papers out of tools that may store them.

4. Prepare a feedback comment bank

Write a bank of 20 feedback comments for [type of assignment], covering common strengths and common problems in [criteria]. Each comment should say what the student did, why it matters, and one specific way to improve. Under 40 words each. Encouraging but honest.

Check: personalise each comment before it reaches a student. A comment that does not match the script undermines your feedback.

5. Redesign an assignment for a world with AI

Here is an assignment brief for [course]: [paste]. Suggest three ways to redesign it so it tests the student's own thinking even if they have AI tools. Ideas could include using local data, a short oral explanation, drafts with reflections, or linking to class discussions. For each, say what it would cost me in marking time.

Check: choose what works for your class size. Tell students clearly what AI use is allowed, as our guide How do I use ChatGPT at university without cheating? suggests they ask.

6. Help students revise

Turn these lecture notes into a one-page revision guide for students: the five key ideas, three common mistakes students make on this topic, and ten self-test questions with answers at the end. NOTES: [paste your own notes]

Check: share it only after you have read it. Our student guide, how university students can revise for exams with AI, shows students how to test themselves this way.

7. Handle research and department admin

Turn these rough notes from our department meeting into minutes with decisions, action points, owners and deadlines. Do not add anything that is not in the notes. NOTES: [paste notes without staff or student names]

The same approach works for grant applications. Paste the funder's published criteria and your own draft, and ask AI to list where the draft does not answer a criterion. Then write the fixes yourself. Check the funder's own rules on AI use first.

Check: the person responsible signs off minutes and applications. Your research ideas, data and findings are yours to protect.

What AI should not do for lecturers

  • Mark student work. AI can suggest wording for feedback. The mark and the academic judgement stay with the marker and the moderation process.
  • Decide misconduct. Do not treat an AI detector score as proof that a student cheated. Follow your university's misconduct process and talk to the student.
  • See student data in a public tool. No names, registration numbers, marks, health information or real scripts.
  • Write your research for you. AI can help with admin and editing. Your ideas, analysis and authorship must be honest and follow your university's research integrity rules.
  • Replace your expertise. Check every fact, reference and model answer. If you would not sign it, do not use it.

How FuKazee trains university staff

FuKazee runs AI training for universities, colleges and schools in Kenya, for lecturers and the admin teams behind them. Staff practise on lecture preparation, rubrics, feedback, coursework design and registry letters, using synthetic examples so no real student work is pasted into a tool. Each programme includes policy material leadership can adapt. As listed on our education page, FuKazee has delivered AI training at the University of Nairobi, and ScholarShift Accelerator cohorts have drawn postgraduate researchers from several Kenyan and international universities. See AI training for universities, colleges and schools, or read what is safe and what works for teachers.

Questions people ask

Is it allowed for lecturers to use AI?

That depends on your university's policy. In general, using AI for preparation and admin, while you check everything and protect student data, is a sensible place to start. Ask your department if you are unsure.

Can I use AI to mark scripts?

Not to set marks. AI can help you write a rubric and draft feedback comments, but the academic judgement must be yours. Pasting real scripts into a public tool also shares student work without permission.

Should I use AI detectors on student work?

Treat any detector result as a reason to look closer, not as proof. Talk to the student, look at drafts and notes, and follow your university's process. An honest conversation is fairer than a percentage.

How do I tell students what AI use is allowed?

Write it into every assignment brief: what is allowed, what is not, and how students should say when they used AI. Clear rules are fairer than a general warning.

Where should a lecturer new to AI start?

Start with one low-risk task, such as a feedback comment bank or meeting minutes. Use a free tool, check every answer, and build from there.

Sources

All sources last checked 15 November 2026.

Where FuKazee fits in

FuKazee is a Nairobi training company that has trained more than 3,000 people since 2022. We train lecturers and university staff to use AI on real teaching and admin work, with clear integrity and data rules. Tell us which faculties or departments you want to train and what you want AI to help with. Email lead@fukazee.com

Want guided support instead?

Not sure which programme fits? Tell us who you are and we will point you to the right path: corporate AI training, the ScholarShift Accelerator for postgraduates, or KCSE study plans and tuition.

Rodney MainaFounder & Director, FuKazee

Rodney has run FuKazee's training since 2022 and still teaches inside every programme line: corporate sessions, the postgraduate cohort and the KCSE examiner webinars. First Class Honours in Journalism and Mass Communication, University of Nairobi.