The African Development Bank (AfDB) and the United Nations Development Programme (UNDP) have announced plans to accelerate an initiative to mobilise $10 billion for artificial intelligence in Africa, with the goal of helping to create 25 million jobs by 2035.
The potential economic impact is enormous. However, the jobs and wealth that AI can generate are not limited to sectors that use the technology; they also lie in the industry that makes AI possible.
Africa cannot afford to remain merely a customer in those markets.
With so much attention focused on ChatGPT, it is easy to overlook the fact that some of the biggest fortunes in AI are made far from conversations with chatbots.
Nvidia, which supplies the chips essential to large AI models, generated $96.2 billion in revenue and $59.7 billion in net profit in the quarter ended July 2026 alone, according to Louis-Nino Kansoun in an article published by Ecofin Agency.
OpenAI, whose applications dominate much of the public conversation, had reached an annualised revenue run rate of nearly $50 billion by late September, according to Reuters. Although the figures cover different periods and measure different things, the contrast is striking.
Nvidia is also just one part of a vast value chain. Its chips are manufactured in part by Taiwan’s TSMC, which generated $122.4 billion in revenue in 2025.
Dutch company ASML, which supplies equipment for manufacturing advanced semiconductors, recorded €32.7 billion in sales.
The computing power behind AI also requires infrastructure to operate. Amazon Web Services generated $128.7 billion in cloud revenue, while U.S.-based Vertiv, which supplies power and cooling equipment for data centres, reported $10.2 billion in sales.
Beyond these companies lie markets for data, AI models and enterprise software. While not all of this revenue comes exclusively from AI, the figures illustrate the scale of the industrial and digital economy that its growth supports.
The idea that Africa can secure a place in this industry is far from unrealistic. The continent does not need to build the extremely expensive factories that manufacture the world’s most advanced chips in order to participate.
It can compete in semiconductor assembly, testing and electronic design.
In Morocco, STMicroelectronics already assembles and tests semiconductors for industries including automotive manufacturing at its Bouskoura plant, which employs about 3,100 people.
In Egypt, authorities decided in May 2026 to extend export incentives to electronic design, semiconductors and embedded systems.
Not every African country can compete in the same activities, but that should not prevent countries from identifying the parts of the industry where they can become competitive.
There is little value in developing one national AI strategy after another if those strategies do not also help create businesses capable of selling products and services to the global AI industry.
For economies seeking new sources of growth, AI represents far more than a technology to adopt. It offers markets to compete for.
Digital sovereignty matters, but it is no substitute for industrial policy.
Africa does not need to manufacture everything, but it must build the capacity to produce what this new industry requires.