Kenya’s AI skills gap is real and the government says so in writing: its own Kenya Artificial Intelligence Strategy 2025-2030 describes “a critical gap in the labour market, with demand for ICT and AI skills far exceeding supply.” The gap has two layers. At the top, there are too few people who can build and train AI models. Across ordinary workplaces, there are too few staff who can use AI tools well, check their output and keep AI systems running after the vendor leaves.
The good news is that the second layer, the one costing most businesses money, is the fastest to close. It does not need a degree. It needs targeted practice on the work people already do.
Key Takeaways
- Treat the gap as two problems: a shortage of AI builders (slow to fix) and a shortage of confident AI users and system owners (fast to fix with practice)
- Use the free programmes from government, Microsoft, AWS and Moringa for foundations, but do not expect a free online course alone to change how your team works
- Train the managers who approve AI spending first, because a manager who cannot judge an AI proposal will buy the wrong thing
- Tie every training session to one task the person will do differently on Monday, then protect two weeks of practice time
- Teach people to check AI output, not only to produce it; this is the skill most often missing
How big is the AI skills gap in Kenya in 2026?
There is no single national number for “missing AI workers”, so be wary of any article that quotes one. What we do have are three credible signals that point the same way.
The government’s own diagnosis. The national AI strategy, launched in March 2025, says Kenya “faces a huge shortage of advanced-level AI experts with the skills to design and train contextually relevant AI models”, and that this forces “reliance on foreign expertise.” It also reports “mismatches between academic training and practical industry needs.” Citing the Oxford Insights readiness index, it notes that only 25% of university graduates complete a STEM course. The strategy even flags, in its own margin note, that the definition of the gap should be expanded “to include non tech professionals”. That is the layer most businesses feel.
What employers report. The Moringa Kenya Technology Workforce Capability Report 2026, based on interviews with 166 senior executives across eight sectors, found that 48% named AI engineering as the hardest capability to build internally, rising to 62% in large enterprises. Some 73% of organisations placed themselves at an early or developing stage of tech workforce maturity. Moringa is a training provider, so read it as one informed view rather than a neutral census.
What regional CEOs are doing. KPMG lead economist Frank Blackmore told Capital FM in November 2025 that 40% of East African CEOs are already investing in AI and 62% are actively recruiting tech-savvy talent, and warned that uneven access to skilled talent could slow AI growth.
Our view: the headline shortage of PhD-level AI engineers is a national problem that universities and policy will take years to fix. The shortage that hurts a Thika distributor or a Kisumu SACCO this quarter is simpler. Nobody on the team knows how to brief, check or maintain the AI tools they already pay for.
Which AI skills are Kenyan employers actually short of?
It helps to split AI skills into three tiers, because each one has a different fix and a different timeline. If you want the employer side in more detail, our guide to the AI skills Kenyan employers want goes role by role.
| Skills tier | Who needs it | What it looks like at work | How fast it can be built | What this means for you |
|---|---|---|---|---|
| AI literacy | Every staff member, especially managers and anyone approving spending | Knowing what AI can and cannot do, writing clear prompts, spotting wrong or invented output, handling customer data carefully | Days to a few weeks of guided practice | The cheapest win. Start here before buying any new tool |
| AI practitioner (system owner) | The people who run your chatbot, automation or dashboard day to day | Updating a chatbot’s answers, fixing a broken automation, keeping data clean, reading a forecast’s confidence | Several weeks, with supervised use on your real system | This tier ends the “call the vendor for every change” habit |
| AI specialist (builder) | Technical staff building custom models or integrations | Data engineering, model training and evaluation, API integration | Months to years | Hard to hire and hard to keep. Most SMEs should rent this skill, not build it |
Most Kenyan SMEs need tier one for everyone and tier two for two or three people. Very few need tier three in-house. Large organisations, as the Moringa data suggests, feel the tier three pain most.
What is the Kenyan government doing to close the AI skills gap?
Talent development is Goal 4 of the national AI strategy: “Build a robust pipeline of competent and agile AI workers for Kenya.” Its flagship projects include AI and data science in school curricula at all levels, a training of trainers programme for AI educators, common AI courses in tertiary education, mentorship and apprenticeship programmes, extending the Presidential Digital Talent Programme (PDTP) and Public Service Internship Programme to allow specialisation in AI and data analytics, mapping AI talent for gaps, and a national programme to attract foreign AI talent, including a review of work visa policy.
Alongside the strategy, several public and private programmes are running. Here is what we could verify, and who each one actually serves.
| Programme | What was announced or reported | Who it serves | What this means for you |
|---|---|---|---|
| Microsoft AI skilling initiative | Announced in November 2024 with the Ministry of ICT: train 1 million Kenyans in AI and cybersecurity by 2027, including 300,000 public servants over three years | SMEs, professionals, educators, students, public servants | Large reach, but we found no published progress figure, so check what is open to you now |
| Microsoft Elevate AI National Skilling Initiative | Launched with Stanbic Kenya Foundation, Pathways Technologies and Konza Technopolis; 152 participants trained by June 2026, with content on agriculture and small business | Trainers and institutions, aimed at underserved and informal sectors | Useful if you run a training institution or co-op, early days in scale |
| Moringa with Google.org | A $2 million grant to train 3,600 Kenyan youth in AI for free by December 2026, starting with software engineering | Young people, with a focus on women and underrepresented groups | A strong route for a young developer, not for your sales team |
| AWS AI Ready | Free online courses, labs and webinars, from foundational courses for non-technical executives to advanced ones | Anyone with internet access | A free starting point for managers who want the basics |
| JKUAT, Samsung and UNESCO | The Standard reported in October 2026 80 JKUAT students on a three-month AI and coding programme, and over 200 TVET teachers upskilled through UNESCO’s Global Skills Academy | University and TVET students and teachers | Builds the future pipeline; does little for staff already on your payroll |
Our honest read: the public effort is real, but it leans towards youth, students, teachers and public servants. The mid-career accountant in Nakuru, the shop manager in Eldoret and the operations lead in a Mombasa logistics firm are mostly left to their employers. That is where the gap will persist longest.
Why don’t free online AI courses close the gap on their own?
Because finishing a course and changing how you work are different things. The same Moringa report found that 77% of organisations use online platforms for staff upskilling, yet only 16% rate their current approach as very effective. Employers said hands-on learning, training aligned to business needs and flexible schedules were what made the difference.
Moringa’s CEO put it plainly: “Technology skills create value when people can apply them to real problems.” Speaking at a Moringa graduation reported by The Standard, the ICT Principal Secretary framed industry-led skills training as a critical pillar of the AI strategy.
We agree. A certificate proves someone watched the videos. It does not prove they can turn a messy month of M-Pesa statements into a clean summary, or notice when the AI quietly made up a number. We learned that second lesson on our own publishing desk: an AI writer will invent statistics and client stories unless checks stop it. Checking output is a trainable skill, and it is the one most courses skip.
How can a Kenyan employer close the AI skills gap? A plan for this week
These steps work for a 10-person shop or a 400-person firm. The difference is only how many people sit in each group.
- List the AI tools you already pay for or plan to buy. ChatGPT or Copilot licences, a WhatsApp chatbot, a forecasting dashboard. Training starts from real tools, not from “AI” in general.
- Name the owner of each tool. One person per tool who will be trained to tier two. If nobody owns it, nobody will maintain it.
- Run a 15-question skills check. Ask each person what they can do today: write a usable prompt, spot an invented fact, update the chatbot, explain a forecast. It is a map, not an exam.
- Train managers first. A short, example-led session for anyone who approves AI spending. Our guide on what Kenyan team leaders should master first lists the essentials.
- Pick one task per person. Drafting supplier emails, summarising sales reports, answering common customer questions. Training is built around that task and tested on it.
- Protect two weeks of practice time. Adjust workloads so people use the skill daily. Without practice, what people learn in a workshop fades.
- Set your data rules. Decide what must never go into a public AI tool, such as customer ID numbers or health records, in line with the Data Protection Act 2019.
- Review at 90 days. Is the tool used more? Are vendor calls down? Are overrides of AI recommendations written down with reasons? Retrain where the numbers say so.
For format choices (in-house, cohort, online), our guide to corporate AI training programmes in Kenya compares the options.
Worked example (hypothetical): a 45-staff building materials distributor in Thika has paid KSh 1,500,000 for a stock forecasting system. Assume the team uses it for only half of what it was built to do because managers do not trust its recommendations and data entry is inconsistent. In effect, 1,500,000 × 0.5 = KSh 750,000 of that investment is sitting idle. Training the three managers to read the forecasts, the two data clerks on clean entry and one administrator on the monthly update targets exactly that idle value. The caveat: if the system itself is badly fitted to the business, training will not rescue it, and you should find that out in step 3.
How can an individual in Kenya build AI skills without a tech degree?
You do not need coding or advanced maths for tiers one and two. Here is a sensible order.
- Start with a free foundation. Elements of AI, from the University of Helsinki and MinnaLearn, is a free online course that needs “no complicated math or programming” and runs at your own pace. AWS AI Ready also offers free foundational courses.
- Practise on your own job. Use a general assistant such as ChatGPT, Gemini, Copilot or Claude on a real task each day, then check every fact it gives you.
- Build proof, not just certificates. Keep three before-and-after examples of work you improved with AI. Employers can judge those in a minute.
- Go deeper only when you know your direction. If you want to become a developer, programmes like Moringa’s are built for that. If you want to stay in your field and be the person who makes AI work there, applied training is a better fit.
Our guide on how to learn AI online in Kenya without a technical background walks through this in more detail, and how to pick an AI course that actually builds skills helps you judge paid options.
Where does paid AI training fit?
Paid training earns its cost when free material has run out of usefulness: when you need practice on your own tools and data, feedback from someone who has seen the common mistakes, and a structure that makes people actually finish.
AI Consultancy Kenya trains both individuals and teams. You can see the current options on our training page. For teams, we discuss your situation first and then quote in writing based on scope.
Mistakes to avoid when closing the AI skills gap
- Training the wrong people first. Budget spent on staff who never touch the tools is wasted. Start with tool owners and decision-makers.
- Generic courses with no application target. If nobody can say what participants will do differently next week, the course is a certificate exercise.
- No practice time after training. Skills fade fast without daily use. Plan the workload, do not just hope.
- Teaching prompting but not checking. Fast output that contains invented facts is worse than slow, correct work.
- Ignoring data protection. Pasting customer records into a public tool is a Data Protection Act risk, whatever the tool’s convenience.
- Trying to hire your way out. Experienced AI specialists are scarce, as both the strategy and employers report. Growing two internal champions is usually more realistic.
Frequently asked questions
What is the AI skills gap in Kenya?
It is the difference between the AI skills employers need and the skills the workforce has. Kenya’s AI strategy describes demand for ICT and AI skills “far exceeding supply”, with the sharpest shortage among advanced AI experts. In everyday businesses, the gap shows up as staff who cannot use, check or maintain AI tools well.
Is the government training Kenyans in AI for free?
Several free programmes exist, often run with partners. Microsoft announced a goal of training 1 million Kenyans in AI and cybersecurity by 2027, Moringa and Google.org are training 3,600 youth by December 2026, and AWS offers free online courses. Eligibility and intake dates change, so check each programme directly before planning around it.
Do I need maths or coding to learn AI in Kenya?
Not for using AI tools well or for running AI systems at work. You need them only if you want to build models yourself. Our guide on AI course requirements in Kenya explains which paths need which background.
How long does it take to close an AI skills gap in a small team?
AI literacy for a small team can be built in days to a few weeks if people practise on real tasks. Training a system owner takes several weeks of supervised use. Building in-house AI engineering talent takes months or years, which is why most SMEs rent that skill instead.
Should we hire an AI specialist or train our existing staff?
For most Kenyan SMEs, train existing staff first. They already know your customers, products and processes, which a new hire would take months to learn. Hire or contract a specialist only when you need custom models or integrations that off-the-shelf tools cannot handle.
The bottom line
Kenya’s AI skills gap is officially recognised and partly funded, but national programmes mostly target students, youth and public servants. For the people already in your business, closing the gap is your job, and it is cheaper than it looks: managers who can judge AI proposals, two or three trained tool owners, and everyone checking AI output before it goes anywhere.
If you would like help working out where your team should start, or you want to build your own AI skills, WhatsApp us on 0711 344 702.