AI on African Terms Tests Zambia's Digital Ambitions

Analysis

Artificial intelligence is forcing a blunt question onto Zambia's digital economy: who owns the systems AI runs on, not simply where the data sits.

That is the argument at the centre of a long-read published by African Business on 6 July 2026. Writing for the pan-African business title, Dr Bulelani Jili argues that AI promises real opportunities for African firms, but risks deepening technological dependency unless African states and companies own more than the location of their data.

Localisation is not sovereignty

Jili draws a sharp line between two ideas that get treated as one. Data localisation means storing information inside national borders. Digital sovereignty is broader: control over the infrastructure, talent and platforms that turn that data into value. The distinction matters because Africa's cloud market remains concentrated among a handful of foreign providers: Amazon Web Services, Microsoft Azure, Google Cloud and Huawei Cloud, according to the analysis.

Real digital sovereignty will depend less on where data is stored than on who owns the infrastructure, talent and platforms that create value from it.

— Dr Bulelani Jili, writing in <a href="https://african.business/2026/07/long-reads/building-ai-on-african-terms-digital-sovereignty-beyond-localisation">African Business</a>, 6 July 2026

Zambia's own infrastructure test case

Zambia does not have to look far for the tension Jili describes. Kwacha News reported in May that the Smart Zambia Institute and Huawei signed a memorandum of understanding for a national AI data centre, AI-powered services across all 25 government ministries, and 5,000 trained ICT professionals by 2028. Vice President Mutale Nalumango officiated at the signing, calling it proof of Zambia’s ambition to become a regional digital hub.

The deal sits inside the wider Digital Zambia Acceleration Programme, which also commits to 2,000 kilometres of new fibre-optic cable, 500 connected government institutions and four million digital identities. Each element expands what Zambia can do online. The arrangement tests whether Zambia can convert new capacity into real ownership, not just access. That is the question Jili puts at the centre of his analysis.

That question is bigger than Zambia. Kwacha News has reported that Africa holds under 1% of global data-centre capacity despite housing about 18% of the world’s population, with more than 70% of global capacity concentrated in North America, Europe and East Asia. Zambia does not yet sit among the four African markets that already host the bulk of the continent’s capacity (South Africa, Nigeria, Kenya and Egypt), which is exactly why the Huawei-backed data centre carries so much weight.

Mobile money: Zambia's application-layer advantage

Jili's essay locates African agency less at the foundations of the AI stack, where compute and frontier models are built, and more at what he calls the application layer: locally adapted services built on top of infrastructure African firms neither own nor govern. Zambia's mobile-money market is the clearest local example.

Kwacha News examined this pattern in Zambia’s mobile-money system, built for the basic handset and the agent kiosk rather than a banking app most users do not have. Airtel Africa and South Africa’s MTN Group run the country’s two largest mobile-money platforms, and neither owns the cloud infrastructure or handset operating systems underneath them. The value Zambia captures sits in the adaptation, not the ownership. That is precisely Jili’s point about the application layer.

The scale of that adaptation is measurable. Mobile-money transactions processed through Zambia's payment system reached 4.07 billion in 2024, up 9.4% on 2023, according to Bank of Zambia payment systems data. Active mobile-money wallets rose to 12,328,755 (84.2% of registered wallets, up from 74.8% a year earlier) after a clean-up of dormant accounts. The Bank of Zambia's National Financial Switch, which now links 14 of the country's 18 commercial banks, is the interoperability layer that lets that volume move between networks at all.

E-government and the data question

Zambia’s other sovereignty lever is legal rather than technical. The Data Protection Act, 2021 came into force on 1 April 2021, establishing the Office of the Data Protection Commissioner to register data controllers and processors and set rules for how personal data is handled. That is exactly the kind of policy agenda Jili’s analysis credits as a genuine sovereignty tool — legislation, not just geography.

The same tension returns once the law meets the infrastructure. The ministries running AI-powered services under the Huawei memorandum, and the four million digital identities the government wants to issue, will generate the kind of sensitive personal data the Act is designed to protect. Enforcement capacity at the Data Protection Commissioner’s office, not the text of the law, will decide whether that protection is real.

What this means for Zambia

The upshot: Zambia has moved furthest on the two legs of digital sovereignty that do not require billions in capital: legislation and application-layer adaptation. The Data Protection Act gives it a legal framework most peers in the region still lack, and its mobile-money market shows genuine local agency. What this means for the harder leg — who owns the compute, the data centre and the cloud — is less settled. The Smart Zambia-Huawei deal answers the question of capacity. It does not yet answer the question of ownership.

Part of Kwacha News’s technology and AI coverage is tracking exactly that gap: where Zambia’s digital ambitions are matched by domestic control, and where they still run through a single foreign vendor’s balance sheet.

What to watch

Three markers will show whether Zambia closes the gap Jili describes. First, how many of the 5,000 ICT professionals promised under the Huawei academy programme are trained and retained in Zambia by the 2028 deadline, rather than the number simply being announced. Second, whether the Data Protection Commissioner’s office publishes enforcement action, not just registration statistics. Third, whether Jili’s promised follow-up analysis, which he says will detail how African governments can negotiate better terms, offers a template Zambia’s own negotiators use.

Frequently Asked Questions

These are the questions readers have been asking since the African Business analysis on AI and digital sovereignty was published. Short answers follow, drawn from that analysis, Bank of Zambia data and Kwacha News's own reporting on Zambia's technology sector.

What is digital sovereignty?

In short, digital sovereignty is control over the infrastructure, talent and platforms that turn data into economic value, not simply where that data is stored. The answer, simply put, is that storing data in Zambia means little if the servers, cloud software and expertise processing it remain foreign-owned. The key is ownership of value creation, not geography.

How does Zambia’s mobile-money system show AI-era digital agency?

Zambia’s mobile-money operators built menu-based services for the basic handsets and agent networks people already use, rather than waiting for smartphone banking apps to spread. Data from the Bank of Zambia shows 4.07 billion transactions moved through the system in 2024. Research on the continent’s fintech sector reveals this application-layer adaptation as the clearest form of local agency, even while the cloud and telecoms infrastructure underneath remains foreign-owned.

Why is the Smart Zambia-Huawei deal different from ordinary digital sovereignty policy?

Ordinary policy, like a data protection law, sets rules Zambia controls. According to the terms Smart Zambia and Huawei announced at the 2026 Zambia Mobile Congress, the national AI data centre, its ministry-wide services and its ICT training pipeline all run through one foreign vendor. The answer is that the deal builds capacity fast, but it does not by itself build ownership.

Who benefits first from AI on African terms in Zambia?

Mobile-money operators, government ministries adopting AI-powered services, and the ICT graduates of the Huawei training programme benefit first. In other words, the near-term winners sit at the application and skills layer. Analysis of the underlying infrastructure shows the harder, more durable question remains open: who owns the compute Zambia’s AI runs on.

What are the real risks of building AI on foreign infrastructure?

Evidence from Jili’s analysis reveals a specific risk: African economies can supply data, labour and consumers to the AI economy while higher-value ownership stays offshore. For Zambia, that shows up in three places: compute billed in foreign currency, cloud contracts governed by foreign law, and a training pipeline run by the same vendor building the infrastructure. Each risk concerns the terms of participation, not exclusion from AI altogether.

Sources

African Business: Building AI on African Terms: Digital Sovereignty Beyond Localisation, by Dr Bulelani Jili, 6 July 2026. Bank of Zambia: payment systems statistics. Data Protection Act, 2021: full text via ZambiaLII. Office of the Data Protection Commissioner: dataprotection.gov.zm. Smart Zambia Institute: szi.gov.zm.