AI is forcing education institutions to reconsider not only what they teach, but how they teach it.
NAIROBI, Kenya, Aug 27 — Artificial intelligence is already changing how Kenyan students learn, even as schools and universities race to determine how they should teach in an AI-enabled world.
Students are using generative AI to explain difficult concepts, prepare presentations and organise assignments, while teachers are experimenting with the technology to develop lesson plans, generate quizzes and adapt explanations.
The technology is forcing education institutions to reconsider not only what they teach, but how they teach it and how they determine whether students have genuinely learned.
A 2025 study involving 137 educators from thirteen Kenyan universities found that lecturers saw significant potential in generative AI but also expressed concerns about academic dishonesty, data privacy, accuracy and over-reliance on the technology.
Educators identified literature searches, summarising academic material, preparing teaching resources and generating ideas among the areas where AI could be useful.
The debate comes as Kenya expands digital access in schools.
The Ministry of Education said in July that 10,382 laptops and an equal number of interactive smart boards were being distributed to Junior Schools across the 47 counties under the Kenya Digital Economy Acceleration Project. The programme followed training for more than 62,000 teachers on ICT integration.
A UNICEF-supported programme aimed at connecting 1,000 public primary schools to reliable internet had connected 363 schools by December 2025, reaching more than 277,000 learners. More than 4,400 teachers had begun integrating digital tools into classroom instruction.
The disparity means the question is no longer simply whether AI should enter Kenyan classrooms, but who will have meaningful access to it and whether teachers are equipped to use it responsibly.
UNESCO Director-General Audrey Azoulay has cautioned that technology should complement rather than replace teachers.
“Digital technology in schools should not be an end in itself, but a means to an end,” Azoulay said, adding that “no screen can replace a teacher.”
Her warning is particularly relevant in Kenya, where access to devices, connectivity and digitally trained teachers remains uneven.
Perhaps the biggest disruption caused by generative AI will not be inside the classroom, but in how learning is assessed.
For decades, schools and universities have relied on essays, take-home assignments, research projects and examinations to establish what students know. Generative AI has complicated that model.
A learner can now produce an essay, summarise a book, prepare a presentation or obtain assistance with some mathematical problems within seconds.
The challenge for educators is therefore no longer simply preventing students from using AI. It is designing assessments that still require them to demonstrate genuine understanding.
A paper published in the Journal of the Kenya National Commission for UNESCO examined the potential of generative AI in developing examination test items, noting that AI could help assessment bodies generate large numbers of questions but still requires subject-matter experts to ensure quality and appropriate cognitive levels.
Kenya’s 2025 National Qualifications Framework regulations require assessment methods to be guided by expected learning outcomes and Bloom’s Taxonomy, while recognising different approaches to measuring learning achievement.
The shift does not necessarily mean abandoning traditional examinations. Instead, institutions may need greater emphasis on assessments in which students explain their reasoning, defend conclusions, demonstrate practical skills and apply knowledge to unfamiliar situations.
Some Kenyan universities are already building capacity around the issue. The Centre for Teaching and Learning Excellence at the Kenya College of Accountancy (KCA) University held a workshop in February 2025 to train educators on responsible use of generative AI in instruction, research and assessment.
At Kenyatta University, a 2025 workshop brought together participants from nine universities to explore generative AI in higher education teaching, learning, research and assessment.
The challenge now is translating such discussions into assessment practices that measure what learners actually know and can do.
Much of the debate has focused on whether students should be allowed to use AI.
But an equally important question is whether teachers themselves are prepared to operate in an AI-enabled education system.
A teacher unfamiliar with generative AI may struggle to identify AI-assisted work or recognise when an AI-generated answer is inaccurate. Similarly, a lecturer who continues assigning work that a chatbot can complete in seconds may inadvertently reward access to technology rather than genuine understanding.
Moses Gitonga, a part-time lecturer at Kenyatta University, said AI should be treated as an assistant to the teacher rather than a replacement for professional judgement.
“A teacher can use AI to generate lesson ideas, develop questions, simplify difficult concepts and even suggest different ways of presenting a topic. But what comes from AI is only a starting point,” Gitonga said.
“The teacher must interrogate that information, verify it and determine whether it is appropriate for the learners before using it.”
He said educators must also understand AI’s limitations, particularly its ability to generate convincing but inaccurate information.
“The biggest mistake would be to assume that because an answer is well written, it is necessarily correct,” he said, adding that teachers must fact-check AI-generated material, identify bias and avoid entering learners’ personal information into such systems.
Teacher training will therefore be critical.
More than 62,000 teachers had received ICT integration training by July 2026, according to the Ministry of Education. But AI literacy requires more than knowing how to operate a device.
Teachers will need to understand hallucinations, bias, privacy, copyright, academic integrity and the circumstances in which AI should not be used.
Universities face an even broader challenge: preparing lawyers, journalists, accountants, engineers, doctors, teachers and software developers for professions that are themselves being transformed by AI.
The question is no longer whether graduates will encounter AI in the workplace. It is whether they will know how to use it without surrendering professional judgement.
A law student may need to verify AI-assisted legal research. A journalism student may have to authenticate synthetic photographs, video or audio. A medical student may need to understand both the potential and limitations of AI-supported diagnosis.
Kenyan universities are beginning to incorporate these questions into academic discussions, including through workshops examining AI’s implications for teaching, research and professional practice.
This does not mean every graduate needs to become a programmer.
Instead, graduates will need enough AI literacy to understand what these systems can do, where they can fail and when human expertise must take precedence.
AI literacy is increasingly becoming a professional competency rather than simply a technology skill.
Perhaps the biggest misconception about preparing young people for an AI-driven economy is that the answer is simply to teach more coding.
Technical skills will remain important, but AI is also likely to increase the value of capabilities requiring human judgement.
The World Economic Forum’s Future of Jobs Report 2025, based on the views of more than 1,000 employers representing more than 14 million workers across 55 economies, found that analytical thinking remained the most sought-after core skill.
Resilience, flexibility and agility, leadership and social influence and creative thinking also ranked highly, while AI and big data, networks and cybersecurity and technological literacy were among the fastest-growing skills.
The report estimates that 39 per cent of workers’ existing core skills could change by 2030, underscoring the need for continuous learning and retraining.
For Kenya, AI literacy should therefore become an extension of digital literacy.
Learners need to know how to use AI tools, but also how to question their answers, verify information, identify bias, protect personal data and recognise when technology should support rather than replace their thinking.
They will also need communication, collaboration, creativity, problem-solving and the ability to learn continuously.
The advantage may belong to those who know when to use AI, how to question it and when not to trust it.
Kenya’s Competency-Based Curriculum already places emphasis on critical thinking, creativity, collaboration and digital literacy.
The challenge is ensuring those competencies are delivered consistently across the education system.
The government’s expansion of digital devices is an important step. The distribution of 10,382 laptops and smart boards is intended to expand access across Junior Schools, while internet connectivity programmes are attempting to address another major barrier.
But infrastructure alone will not determine whether AI improves education.
A learner in a well-resourced school with reliable internet, devices and digitally trained teachers may have a very different experience from one in a school where connectivity remains unreliable.
Research by Rest of World in 2025 highlighted the uneven nature of AI adoption in Kenyan schools, reporting that teachers were already using ChatGPT and other AI tools to manage workloads while limited internet connectivity remained a major obstacle to scaling AI-based education.
If access to AI becomes another educational advantage available mainly to better-resourced learners, the technology could widen existing inequalities.
But if deployed responsibly, it could also provide additional learning support, help teachers prepare materials and give students access to information and personalised assistance.
The difference will depend on infrastructure, teacher capacity, curriculum design and clear rules governing how the technology is used.
The education debate should therefore move beyond whether AI is good or bad.
Schools cannot realistically prepare learners for the future by pretending the technology does not exist. Nor should they allow students to outsource learning and independent thinking to machines.
The more useful approach is to teach students how to work with AI critically and responsibly.
That means redesigning assessments where necessary, training teachers, updating curricula and giving learners practical experience with technology while maintaining human accountability.
It also means teaching students that an AI-generated answer is a starting point, not necessarily the final answer.
For Kenya, the goal should be to produce learners who can use technology without becoming dependent on it, recognise its strengths without overlooking its weaknesses and continue learning as the technology evolves.
Artificial intelligence may reshape the workplace, but education will determine whether Kenyan learners enter that future as creators of technology, informed users of it — or merely spectators watching others build the future.
