--- title: "Who Will Control Africa's AI Infrastructure, and at What Cost?" description: "A wave of foreign investment is wiring Africa for the AI age — undersea cables, data centers and GPU clusters. As US tech giants, Gulf funds and Chinese firms compete for position, African policymakers are asking a harder question: who owns the infrastructure, and who decides how it is used?" category: "Technology" category_url: https://newsparlor.com/category/technology author: "Noah Andersen" published: 2026-06-26T02:12:48.000Z updated: 2026-06-26T02:12:48.000Z canonical: https://newsparlor.com/article/who-will-control-africa-s-ai-infrastructure-and-at-what-cost tags: ["Africa", "artificial intelligence", "data centers", "digital sovereignty", "technology"] --- # Who Will Control Africa's AI Infrastructure, and at What Cost? A wave of foreign investment is wiring Africa for the AI age — undersea cables, data centers and GPU clusters. As US tech giants, Gulf funds and Chinese firms compete for position, African policymakers are asking a harder question: who owns the infrastructure, and who decides how it is used? Africa is home to roughly 18% of the world's people but holds less than 1% of its data center capacity. That gap is now drawing some of the largest technology investments in the continent's history — and with them, a debate about who will own the infrastructure, who will profit, and who will bear the costs. ## A continent being wired The incoming capital is substantial. Google's Equiano subsea cable runs along Africa's western coast, and Meta says the core of its 45,000-km "2Africa" cable, touching dozens of countries, was completed in late 2025, [the company reported](https://engineering.fb.com/2025/11/17/connectivity/core-2africa-system-completion-future-connectivity/). Microsoft and the Abu Dhabi-based AI firm G42 have proposed a roughly $1 billion data center in Kenya — still under negotiation as of mid-2026 — paired with local-language AI work and skills training, [Al Jazeera reported](https://www.aljazeera.com/news/2026/6/26/who-will-control-africas-ai-infrastructure-and-at-what-cost). Nvidia-linked "AI factory" projects have been announced in South Africa. ## A multipolar contest The competition is genuinely multi-sided. US hyperscalers — Microsoft, Google, Meta, Amazon — are racing for cloud regions and cable landing rights. The UAE's G42, backed by Abu Dhabi capital, has positioned itself as a bridge investor. Chinese firms retain a large footprint in terrestrial networks. African capital is entering too, through continental funds and a new Africa AI Council set up to coordinate strategy. Some analysts argue the rivalry can work in Africa's favor. Priyal Singh of Signal Risk suggested that African states will get "greater room for manoeuvre on AI and data infrastructure, precisely due to how contested and fragmented this industry is amongst global leaders," [Al Jazeera reported](https://www.aljazeera.com/news/2026/6/26/who-will-control-africas-ai-infrastructure-and-at-what-cost). ## The upside: languages, jobs, services Supporters point to concrete benefits. Investment promises connectivity, jobs and training, and a chance to build AI that actually serves African users. Homegrown startups are already doing so: South Africa's Lelapa AI has built a multilingual model covering languages such as Swahili, Yoruba and isiZulu, and other African firms and governments have launched local-language voice and text systems that big English-first models have long underserved, [Rest of World reported](https://restofworld.org/2026/africa-ai-sovereignty-big-tech/). Tools in agriculture, health and public services are seen as especially promising. ## The sovereignty question The concerns are equally concrete. Policy analysts have long warned that a large majority of African data is processed or controlled by foreign entities — figures often cited at 85–95%, though the underlying methodology is hard to verify — raising fears of dependence and lost value. Several governments have recently rejected data-sharing deals that would have moved citizens' data offshore. Experts frame the dilemma carefully. Sanusha Naidu of the Institute for Global Dialogue said that "whether it's a US company, a company from Europe, or a Chinese company, policymakers must weigh the broader developmental impact." Rachel Adams of the Global Center on AI Governance cautioned that "Africa's push for digital sovereignty cannot mean total independence from global AI supply chains," but argued it must involve enforceable data-control commitments and real local investment rather than promises. ## Power and water AI infrastructure is also resource-hungry. Data centers consume large amounts of electricity and, often, water for cooling — a strain in places where grids are unreliable and water is scarce. Critics say those costs are rarely foregrounded in glossy investment announcements, and that diesel backup for power-short grids undercuts the "green" branding some projects carry. ## Policy racing to catch up Policy is advancing, if unevenly: most African countries either have or are drafting national AI strategies, and many have endorsed a continental AI declaration. But dealmaking is moving faster than rule-making. Joseph Asunka, who heads the Afrobarometer survey network, warned that the stakes are too high for closed-door deals: "These negotiations should not just be conducted at the elite level and dumped on citizens." Whether the build-out becomes a genuine development partnership or a new form of dependency, analysts say, will hinge on what African governments negotiate now — and how openly. ## Sources - [Who will control Africa's AI infrastructure and at what cost?](https://www.aljazeera.com/news/2026/6/26/who-will-control-africas-ai-infrastructure-and-at-what-cost) - [Pushing back from Big Tech: Africa's hard road to AI sovereignty](https://restofworld.org/2026/africa-ai-sovereignty-big-tech/) - [Announcing the completion of the core 2Africa system](https://engineering.fb.com/2025/11/17/connectivity/core-2africa-system-completion-future-connectivity/)