Saturday, November 21, 2026

Subscribe

EconomyEdition ZambiaZambia

Zambia’s AI computing centre must build lasting local capacity

Lusaka is exploring an AI computing infrastructure with Beijing. Its value will also depend on available power, local skills, financing and strong data safeguards.

Travellers and a technician move through a busy bus station in Lusaka, Zambia.

A data centre alone cannot create useful artificial intelligence. It also requires electricity, specialists, data protection rules and applications that meet the country’s needs. Zambia is seeking to address this equation by strengthening its technological cooperation with China.

A project building on existing cooperation

The initiative is not starting from scratch. In May 2026, Smart Zambia, the public body responsible for digital transformation, and Huawei signed a memorandum of understanding on co-investment in national AI and cloud infrastructure. State broadcaster ZNBC also reported that the partnership included training 5,000 Zambian information technology professionals by 2028.

On 9 November, Technology and Science Minister Felix Mutati and China’s ambassador, Han Jing, discussed the next steps in this cooperation. The Chinese embassy’s account confirms the meeting and the minister’s expressed interest in China’s experience in artificial intelligence. Zambian outlet Efficacy News reports that discussions also covered an AI-ready data centre and the development of a roadmap. The two accounts do not detail the same points, however, so the project’s technical and financial arrangements remain to be specified.

An AI-designed centre differs from a facility intended mainly for storage. In particular, it must provide substantial computing power, using specialised equipment capable of training and running models. According to Efficacy News, Mutati said Zambia’s existing infrastructure mainly meets storage needs and is insufficient for advanced applications.

Applications expected in mining and agriculture

The government is highlighting two sectors that are essential to Zambia’s economy: mining and agriculture. In mining, AI could help analyse geological data and guide exploration work. Mutati cited the case of a copper deposit that was reportedly identified more quickly with AI-assisted tools in a cobalt-mining area. Efficacy News reported this account, but it has not been independently confirmed in the sources consulted.

In the fields, analytical tools could help with crop selection, sowing plans and better resource allocation. The government is targeting annual production of 10 million tonnes of maize. AI can support planning, but it cannot by itself guarantee that this target will be met: results will also depend on agricultural conditions, infrastructure and producers’ access to the tools.

These benefits will not be distributed automatically. Large mining companies and government agencies often have more data and resources than small farms or informal businesses. For the infrastructure to serve more than major players, it will be necessary to define who can access computing power, at what cost and on what terms.

Data control and material requirements

Hosting data in the country can reduce reliance on processing carried out abroad. But location alone does not guarantee security or national control. Zambia already has a data protection law, adopted in 2021. The forthcoming national AI framework will need to specify how these rules apply to automated systems, particularly when personal or sensitive data are used.

The ministry also acknowledges the need for safeguards. Its permanent secretary, Brilliant Habeenzu, called for national AI guidelines to be updated and aligned with cybersecurity and data protection rules. According to the OECD, Zambia’s AI strategy, adopted in 2024 for a period running until 2026, is nearing its end. The next phase will need to clarify public responsibilities and oversight mechanisms.

The centre’s operation will also depend on factors less visible than servers. Power supply must be reliable, connectivity adequate and teams able to operate the equipment. The United Nations Development Programme, which has supported computing capacity in six African countries, including Zambia, stresses that infrastructure alone does not deliver results: skills, usable data, clear governance and practical applications are also needed.

Huawei’s selection also raises the question of technological dependence. The partnership could provide equipment, financing and training. However, long-term costs, data access rules, maintenance and skills transfer will need to be made clear. For Lusaka, the challenge is to turn external cooperation into lasting capacity that is operated and controlled locally.

A challenge for Southern Africa

Zambia is part of the Southern African Development Community (SADC), which brings countries in the region together around economic cooperation and development. National computing capacity could serve Zambian government agencies, researchers and businesses. In time, it could also support regional exchanges, subject to agreements on security, costs and data flows.

At continental level, the project illustrates a shared challenge: developing digital infrastructure without limiting ambition to buying equipment. The African Union’s AI strategy calls for the continent’s capacity and skills to be strengthened. For Zambia, the next practical step will therefore be less about announcing a centre than publishing a roadmap: financing, timelines, governance, energy and training will all need to be specified.

Available sources confirm a memorandum of understanding signed in May and bilateral discussions on 9 November. They do not yet establish the cost, construction timetable or technical specifications of the future centre. These details, along with safeguards for sovereignty and access, will determine the project’s real significance.

Automatic translation. Original French version

The quiz

Edition ZambiaZambia in 5 questions

5 questions