Will AI Take African Jobs If We Build The Right Tools?

The founder behind payroll and business systems used across the region offers a grounded answer to the question he hears every few months. The danger, he argues, is less the technology itself and more the habit of building impressive things nobody around us actually needs.

Every few months, someone asks James Ngugi, the Founder & CTO at Attallah Continental Group, whether artificial intelligence is going to take African jobs. He considers it a fair question, and he has settled on an answer that declines both of the usual scripts.

“I try not to answer it with either fear or hype,” he says. “AI is a tool, and like every tool before it, what it does to work depends on what we build with it and who we build it for.”

Coming from a builder whose systems run payroll and business operations across the region, the answer carries a particular weight. Ngugi is neither an outside commentator nor a vendor with a product to defend in the AI gold rush. He is someone who has to decide, in his own roadmaps, what the technology is actually for.

The real risk is quieter than the robots

The loudest fear in the jobs debate is that clever machines will simply replace people. Ngugi’s experience points him at a different failure, one that attracts far less attention.

“The bigger risk is quieter,” he says. “It is that we pour our energy into building impressive AI that solves problems the people around us do not have, while the tools they actually need go unbuilt. A system that dazzles a conference is worth less than a plain one that helps a shop owner get paid. If we chase the impressive over the useful, we lose, and no machine is to blame.”

It is a criticism aimed as much at the builders as at the technology, and it lands harder for being delivered from inside the industry.

The critique has a particular edge in an African context, where engineering talent is scarce relative to the problems available and every team building for applause is a team unavailable for the unglamorous work of getting a shop owner paid. The opportunity cost of the impressive, in other words, is highest exactly where the useful is most needed.

Where AI helps ordinary work

Used well, Ngugi argues, AI expands what a small operation can do rather than shrinking who it employs. He gives working examples rather than scenarios.

“It can help a shopkeeper make sense of her own numbers without hiring an analyst. It can draft the message a young entrepreneur needs to reach a customer. It can take the repetitive parts of a job so a person can spend time on the parts that need judgement,” he says. “That is not replacing the worker. That is giving one person the reach that used to need a team.”

The pattern in those examples is deliberate. In each one, the person stays at the centre of the work and the machine absorbs the drudgery around them. The value lands with a shop owner or a founder, and the capability that once required a payroll of specialists becomes something one person can carry.

Build for the person, not the demo

Ngugi keeps returning to the rule that has governed everything he has shipped, and he sees no reason AI should be exempt from it. Build things people will actually use.

“The test for an AI tool is exactly the test for any tool,” he says. “Does it help someone do real work on an ordinary day, or does it only look good on a stage? Hold ourselves to that, and AI becomes another way to put capability in ordinary hands, which is the whole job.”

He does not wave away the disruption. “I am not naive about what is ahead. Some tasks will change, and people will need support to move with them. But the future of work here will be decided less by the technology and more by our choices about what to build.”

That framing puts the responsibility in an uncomfortable place. If the outcome depends on choices about what to build, then the builders, the investors who fund them and the institutions that train them all hold a share of it, and pointing at the technology becomes a way of avoiding the mirror.

His conclusion turns the original question inside out. “Build tools that expand what people can do, aimed at the people who have been left out, and AI does not have to take African jobs. It can help create the businesses that make them.”

Read Also: The Impact of Artificial Intelligence in Transforming Modern Homes in 2025