The usual debate asks whether AI will destroy jobs. For Kenya, the data points to a different question.
The World Bank’s June 2026 Global Economic Prospects estimates that about 40.0% of jobs in emerging and developing economies are exposed to AI, compared with 60.0% in advanced economies. This is often read as reassurance. We read it differently. Lower exposure reflects an employment base concentrated in low-productivity work: 83.8% of Kenya’s workforce is informal, and formal jobs account for only 16.2%. Kenya is insulated largely because there is less to automate, and the same structure means it captures fewer of the productivity gains.
Adoption is not the constraint where it counts. A World Bank survey in early 2026 found that 42.0% of formal firms in Kenya use some form of AI, against 43.0% in the US. The gap is therefore not between Kenya and the advanced economies, but between formal firms and the informal sector, which that survey does not cover.
The pressure point is the entry level. Only about 105,300 of the 822,100 jobs created in 2025 were formal, so formal entry roles are already scarce. They are also largely routine tasks, which are the most easily automated. Where one experienced employee with AI tools can do the work of several junior hires, an employer has little reason to rebuild the pipeline. Unemployment already stands at 16.8% among those aged 20–24.
Our view is that the risk is not mass redundancy. It is a closed first rung in an economy where the second rung is already narrow. The productivity gains are real, but they accrue to employers and to workers already inside.
We expect the pressure to show up first in graduate recruitment, not in headline unemployment, which means it will be under-measured for some time. (Start your investment journey today with the cytonn MMF, call+2540709101200 or email sales@cytonn.com)














