Artificial intelligence is fuelling a global race for digital infrastructure, but Schneider Electric believes Africa may be asking the wrong question.
Across the continent, governments are focused on generating more electricity to meet growing demand. However, Ifeanyi Odoh, President of Schneider Electric East Africa, argues that increasing power generation alone will not prepare Africa for the AI era. The bigger challenge, he says, is modernising electricity grids that were built for a different time.
“The traditional grid as we know it is already becoming obsolete,” Odoh said. “We need grids that are digitally enabled from the start because the way electricity is generated and consumed has fundamentally changed.”
His remarks come as African countries seek to position themselves for the rapidly growing AI economy. Yet the continent currently has just 360 megawatts of operational data centre capacity spread across 217 facilities in 33 countries, less than one per cent of global capacity. As AI models become more sophisticated and cloud infrastructure expands, electricity demand is expected to rise dramatically, putting additional pressure on power systems that are already struggling in many parts of Africa.
For Schneider Electric, this makes electricity infrastructure the next major battleground in the race for AI investment.
The company says it is positioning itself as more than a supplier of electrical equipment. It wants to become a technology partner helping countries build the digital infrastructure needed for AI. Schneider Electric already provides power infrastructure for more than 38% of Africa’s installed data centre capacity and nearly half of East Africa’s facilities. Across the region, over 80% of data centres rely on some combination of the company’s power management, automation and secure power technologies.
Its message is straightforward: countries that modernise their electricity grids first will be better placed to attract the next wave of AI investments.
The rise of artificial intelligence is also changing the economics of electricity. Traditional enterprise server racks typically consume between 5 and 15 kilowatts of power, while AI server racks require anywhere from 40 to 120 kilowatts. Next-generation systems are expected to approach 200 kilowatts per cabinet. Unlike conventional computing infrastructure, AI systems operate continuously, demanding uninterrupted electricity alongside advanced cooling technologies.
The shift is also changing where electricity is needed. Rather than being located in isolated industrial zones, many new data centres are being built closer to cities to reduce network latency. In Lagos, for example, most data centres are situated within residential and commercial areas. At the same time, electric vehicles are increasing electricity demand as charging infrastructure expands into homes, offices and shopping centres.
“Massive infrastructure investments like data centres are now much closer to residential and commercial developments,” Odoh said. “That changes completely how power needs to be delivered.”
Different African markets are already taking different approaches to supporting the AI economy.
South Africa has emerged as the continent’s leading destination for hyperscale cloud investments following significant improvements in electricity reliability. More than 400 consecutive days without nationwide rolling blackouts, alongside increased private renewable energy generation and power-wheeling reforms, have strengthened investor confidence. These improvements helped attract Microsoft’s R5.4 billion ($300 million) cloud infrastructure expansion announced in 2025.
Kenya, meanwhile, is capitalising on its renewable energy resources. More than 70% of the country’s electricity comes from renewable sources, particularly geothermal power, making it attractive to companies looking to reduce emissions. Some developers are now locating AI data centres close to geothermal plants, including Microsoft’s planned $1 billion AI campus with G42.
Nigeria presents a different opportunity. Although the country has 13.6 gigawatts of installed electricity generation capacity, only about 4.3 gigawatts are consistently available. As a result, many data centres depend on captive gas plants, batteries and other off-grid power systems. Schneider Electric believes Nigeria’s abundant natural gas reserves could support a “gas-to-compute” model, where dedicated gas-powered infrastructure is integrated directly into data centre operations.
Despite their different strategies, each of these markets demonstrates the same reality: reliable electricity has become a critical requirement for attracting AI investment.
Odoh believes Africa’s ageing electricity infrastructure could ultimately become one of its greatest advantages. Rather than spending years upgrading legacy systems, many countries have the opportunity to build modern, digitally managed grids capable of integrating renewable energy, battery storage, distributed generation and AI workloads from the outset.
“Africa has a significant opportunity to rethink how it builds energy infrastructure,” he said.
He believes future electricity networks will function much like intelligent software platforms, constantly balancing power from solar farms, gas plants, batteries, rooftop solar installations and AI facilities in real time.
“We need everything,” Odoh said. “The real question is how to integrate all these different energy sources into one intelligent grid.”
To support that vision, Schneider Electric now provides technologies ranging from medium-voltage power systems and power distribution equipment to liquid-cooling solutions that remove heat directly from AI processors, enabling denser and more efficient computing. The company says it is also increasing local design and assembly of these systems across the continent.
“We’re already able to design these systems and assemble them locally in Africa,” Odoh said. “The digital economy can only become a reality if we solve the underlying energy challenge. Energy is the foundation that makes everything else possible.”