Economics – Industry & Resources · Editorial
By Moakanyi Magazine · Global Issue · June 2026
The world's trade figures are increasingly written in silicon, and Botswana is on the buying side of the ledger. AI-related goods supported global trade and import demand through the period, a reminder that the fastest-moving part of world commerce is hardware the country imports rather than exports. That is not a failing to apologise for; it is a fact to plan around.
AI demand shows up in trade data as servers, chips and the equipment that data centres and networks require. For an economy like Botswana's, the relevant question is not whether it can manufacture that hardware – it cannot, and neither can most countries – but whether it can use the imports to build services and skills that earn their keep. The value contest has moved from making the machine to applying it.
On the buying side of the AI trade
Research tracking the trade has documented how AI-related goods supported import demand across markets. Botswana participates in that flow as a consumer – importing the computing and connectivity that digital services run on, from government systems to the private sector in Gaborone and Francistown. The country is a customer in the AI supply chain, not a supplier, and that position shapes its options.
Being a buyer has a cost: every server and device is an import that weighs on the trade balance and depends on stable foreign exchange managed through the Bank of Botswana. But it also lowers the barrier to using the technology, since Botswana need not build the hardware to deploy the services it enables. The same imports that strain the balance of payments are the inputs to a more productive economy.
You do not have to make the chip to use what it computes.
Turning imports into capability
The opportunity is to convert imported hardware into local capability – digital services, data skills and productivity gains in mining, finance and government. The value is captured not at the point of import but at the point of use, where trained people turn machines into output. A server idle in a data centre is a cost; the same server running a service that improves how Debswana, a bank or a ministry operates is an investment.
That means the AI-goods trade is a prompt for skills policy as much as trade policy. Botswana's return on its imports depends on the people who can run the systems, which makes training and connectivity the real investment behind the hardware. The chip is available to anyone with foreign exchange; the skills to make it pay are what separate the buyers who benefit from the buyers who merely spend.
Imported machines pay off only where local skills meet them.
The cost side worth watching
There is a discipline to importing into the AI boom, because the hardware is not free and the foreign exchange is finite. Botswana benefits most by importing with a purpose – tying the equipment to a service, a productivity gain or a public-sector function that justifies the outlay – rather than acquiring technology for its own sake and watching it depreciate unused.
The same caution applies to connectivity and power, the quiet prerequisites that determine whether imported computing does any work at all. Without reliable bandwidth and electricity, the hardware sits idle and the import becomes a pure cost, which is why the AI-goods opportunity is also an argument for the infrastructure beneath it.
Hardware without power, bandwidth and purpose is a cost, not a capability.
Where the deployment dividend could land
The clearest returns are in the sectors Botswana already runs. In mining, applied computing can improve how Debswana and others plan, sort and forecast; in finance, it can sharpen risk and service in the banks of Gaborone; in government, it can speed the services citizens deal with daily. None of this requires building a chip – it requires deploying imported tools against problems the country already has, which is a far more attainable goal than competing in manufacturing.
That reframes the AI-goods boom from a story about who makes the technology to a story about who uses it best. A mid-sized economy that deploys thoughtfully can capture much of the productivity gain while importing all of the hardware, and Botswana's stability and connectivity give it a reasonable starting position. The advantage goes not to the largest buyer but to the most deliberate one.
The deployment race is one a small economy can actually enter.
The so-what for Botswana is to treat the AI-goods boom as an import to be used well rather than a race it has lost. The country will not out-manufacture the chip makers, but it can out-deploy its peers if it pairs the hardware it buys with the skills, connectivity and purpose to make it productive – turning a line on the import bill into a line on the country's capability, and a cost into a return.
Sources: arXiv
