Kenya’s ambition to join the global artificial intelligence (AI) race risks being derailed by a shortage of skilled talent and limited incentives to attract young people to the tech sub-sector.
A new Technology and Innovation Report by the United Nations Conference on Trade and Development (UNCTAD) highlights these concerns, citing Kenya and Uganda among the developing countries where data annotators earn some of the lowest wages globally—often as little as $2 (about Sh260) per hour.
According to the report, workers frequently face exploitative conditions, including long hours—up to 10 per day—repetitive tasks, and minimal opportunities for career advancement.
“In the data preparation stage, employment can involve exploitative, often precarious working conditions,” UNCTAD notes.
“Data annotators in developing countries often experience difficult conditions…with limited opportunities for career advancement, for example in Kenya and Uganda.”
Risk of diskilling
The UN body further flags the risk of “deskilling,” particularly where highly educated individuals—some with graduate or STEM degrees—are relegated to low-skill microtasks such as image annotation or content moderation.
“In India and Kenya, for example, a survey on microtask platforms and BPO companies showed that highly educated workers were often assigned relatively low-skill tasks,” the report states.
“Such significant wastes of human capital may be exacerbated in increasingly connected job markets, in which tasks are outsourced globally.”
Ken Okolo, Head of Commercial Partnerships for the Middle East and Africa at the UK-based Raspberry Pi Foundation, says the problem goes beyond a skills gap—it’s also about finding the right kind of expertise.
“A key challenge is getting the right talent—people who have properly studied AI—so that we ensure models are trained effectively to perform their intended tasks,” says Mr Okolo.
Other barriers
The lack of skilled talent adds to a host of other barriers that have slowed AI uptake in Kenya. These include data scarcity, underdeveloped tech infrastructure, and public apathy towards AI applications.
One persistent challenge is the lack of locally aggregated data, which forces developers to rely on datasets from other regions—an approach that can introduce biases and limit the relevance of AI solutions to the Kenyan context.
Experts have called on Kenyans to digitise key aspects of daily life to build a local data bank that supports the creation of context-specific AI models.
In response, Kenya has taken legislative steps, enacting the Data Protection Act in 2019 to regulate the processing and use of personal data and establish a licensing framework for data handlers.
Pivoting
Despite these challenges, some corporate firms that have adopted AI report increased operational efficiency, suggesting that returns—though limited—are possible.
In an effort to scale this success, the Kenyan government recently unveiled a national AI deployment strategy focused on healthcare, education, and agriculture.
The plan is built on three pillars: modernising digital infrastructure, creating a robust data ecosystem, and incentivising the development of localised AI models and solutions.
Whether this strategy will be enough to overcome the deep-rooted challenges remains to be seen—but the path to an inclusive AI future in Kenya hinges heavily on improving both skills development and workforce incentives.
