Artificial intelligence is no longer a technology waiting to transform the world. It is already changing how we work, communicate, learn, create and make decisions. Yet universities, which are expected to prepare society for this transformation, appear to be moving too slowly.

In Kenya, this should be a serious concern. Young people are entering a labour market being reshaped by AI, while much of university education remains rooted in curricula and teaching approaches designed for a very different economy of the 80’s and 90’s.

The question is simple: what exactly are Kenya’s universities doing to prepare the next generation for a world in which AI will be embedded in almost everything they do?

Are universities researching how AI can improve healthcare, agriculture, manufacturing, education, journalism, finance, public services and many more? Are they developing solutions for Kenya and other economies? Are they redesigning curricula around the skills employers will need tomorrow? And are they teaching students how to use AI without surrendering their ability to think?

These questions matter because the labour market is changing.

The International Labour Organization says 6.1 per cent of jobs held by workers aged 15 to 29 are in occupations highly exposed to AI-related change. Many are administrative roles that have traditionally provided entry points into the workforce. The ILO stresses that exposure does not automatically mean job losses, but technology is changing the tasks young workers perform.

For Kenya, this should be a wake-up call.

We cannot continue producing graduates using curricula designed for an economy that no longer exists and then expect employers to complete the education process.

This is not an argument against theory. A doctor needs medicine, an engineer needs engineering, a lawyer needs law, a journalist needs journalism and an accountant needs accounting. Deep disciplinary knowledge remains essential.

But AI makes that knowledge more valuable, not less. When machines can produce information almost instantly, the human advantage increasingly lies in knowing what questions to ask, determining whether an answer is credible and deciding what to do with it. That is where curriculum reform becomes urgent.

Keeping course content current is necessary, but it is not sufficient. Graduates need disciplinary depth, critical thinking, communication, collaboration, problem-solving and the ability to exercise independent judgement.

The OECD’s 2026 evidence also calls for caution. It says current evidence does not establish that AI is responsible for worsening youth labour-market outcomes, with deterioration having begun before the rise of generative AI. But AI is changing jobs and tasks, making relevant skills and education increasingly important.

We should not blame AI for every difficulty young people face when entering the labour market. But we must recognise that the nature of entry-level work is changing.

Some tasks that traditionally served as apprenticeships, such as background research, routine drafting and basic analysis, are among those AI can increasingly perform.

So where does that leave the graduate?

Increasingly, employers need people who can think, analyse, communicate and apply knowledge from the beginning. The expectation is shifting from learning basic tasks on the job to demonstrating judgement and problem-solving earlier.

Research by the Strada Education Foundation reinforces this shift. Its 2026 research found that employers rated critical thinking, communication and collaboration among the most important skills for entry-level workers, while work experience was the most important factor in hiring decisions.

Universities therefore need to rethink the relationship between classroom and workplace.

Students should not spend four years listening to lectures and encounter real-world problems only after graduation. Universities need stronger relationships with industry, meaningful internships, practical projects and opportunities for students to work on genuine problems before leaving campus.

Education rests on two enduring pillars: learning and thinking.

AI can make learning faster. But if used badly, it can also make thinking unnecessary.

A student who asks AI to write every essay, solve every problem, summarise every book and analyse every argument may become efficient at producing answers without developing the intellectual muscles required to reach those answers independently.

That is de-skilling, as abilities developed through practice gradually erode when technology takes over the work.

A graduate who has never learned how to research independently, structure an argument, analyse evidence or work through a difficult problem has not been deskilled by AI. They simply never developed those skills in the first place.

Kenyan universities must confront this risk without retreating from technology. The answer is not to ban AI from classrooms. It is to teach students how to use it while deliberately protecting the human capabilities that technology cannot replace.

Universities must also become places where AI is researched, not merely taught. Kenyan institutions have an opportunity to develop AI applications for agriculture, healthcare, education, local languages, finance, manufacturing and public services. We should not simply consume technologies developed elsewhere. Our universities should help build technologies that respond to our own problems.

The future university cannot simply be a place where young people acquire certificates before looking for work. It must be a place where they learn how to learn, how to think and how to apply knowledge in a rapidly changing world.

AI is already changing the workplace. The question is whether Kenya’s universities will change with it.

If they do not, we risk sending graduates into an AI-driven economy armed with yesterday’s skills and expecting employers and young people themselves to close the gap.

That would be more than an education problem. It would be a failure to prepare a generation of Kenyans for the future it is about to inherit.