The continent's cultural complexity is outpacing the tools designed to serve it. Lily Njagi, head of HR, Kenya & Uganda at Unilever, has already watched AI eliminate a role she once held, saying the response is not fear, but ownership.
In 2019, Lily Njagi spent months training a bot. The work was detailed and deliberate but, by her own admission, entirely self-defeating. The bot was learning to do her job, and she knew it. For Lily, then a people partner at Unilever, that experience produced clarity.
“AI is here to stay, that's a fact. I remember thinking, ‘I'm training this bot to do my own job.’ But that didn't stop me. I trained the bot as well as I could. Today, the job I was doing in 2019 has been eliminated, and the bot is doing exactly what I trained it to do,” she says.
Her answer to the displacement question is straightforward. AI can absorb transactional work but it cannot replicate is presence. Lily's argument is that as AI takes the repeatable work off the desk, HR professionals should concentrate their energy on those human touchpoints, rather than competing with tools that will always be faster at processing information.
“For me, there is always going to be a space where AI cannot replace people, it can only support and assist.AI will not come and give you a hug when you are bereaved, nor will it come and celebrate with you when you have a baby. Those touchpoints, the moments that really matter, cannot be replaced by AI right now,” she says.
But her sharper concern lies elsewhere. Working across Africa, supporting teams in Kenya, Uganda and beyond, Lily has observed something that the global conversation about AI is largely missing. The tools practitioners are being handed were not built with Africa in mind.
When she puts a global AI tool to work on engagement planning for her East Africa teams, the output reflects none of that. It produces recommendations built on assumptions drawn from data gathered elsewhere, handing them back to her as though geography were interchangeable.
“I support the Uganda business, and the things that are important to Ugandans are different from the things that are important to Kenyans. What they consider valuable and what we consider valuable are two extremely different things. So, I think AI still has a bit to learn when it comes to our cultural nuances. AI chatbots are still struggling with things that are specific to our countries and our region, which are not the same as anywhere else,” she says.
The problem, Lily argues, is not unique to HR. It runs through every function that relies on AI to understand people, from recruitment and communications to customer engagement, learning and development.
But it is in HR where the cost of cultural misreading is most direct. An engagement plan that does not speak to what employees actually care about signals to those employees that the organisation does not know them.
Lily's position is that Africa still has a window to act, to contribute to the training of AI models rather than simply consuming whatever arrives from elsewhere. That is as much a policy and investment question as it is a technology one, but it starts with practitioners being willing to name the gap.
“I am embracing AI. I love it. But it is about remaining ahead of the curve and seeing how to make the most of it, because many jobs will be lost. The question is what we do about it,” she says.
