A New Geography of Cloud
For a decade, Africa’s cloud story was a South Africa story. AWS launched the Cape Town region in 2020 with three availability zones and local presence in Johannesburg. Google followed with a Johannesburg cloud region in early 2024. The Southern African hub absorbed the continent’s early enterprise cloud demand while the rest of Africa relied on cross-border latency and expensive international bandwidth.
That geography is changing in 2026. According to Console Connect’s analysis of Africa’s cloud data centre buildout, eight new data centers launched across Africa in 2026, 24 African nations now host carrier-neutral data centers, and Rwanda and Zimbabwe are expected to add capacity within one year. The Democratic Republic of Congo has added capacity through Open Africa Data Centres and Raxio. The infrastructure map is being redrawn at a speed the industry has not seen before on the continent.
The primary signal of the shift is Microsoft’s Kenya commitment. Microsoft pledged $1 billion in Kenya — in partnership with UAE-based AI specialist G42 — to establish a geothermal-powered data center with 100 MW capacity targeted by 2026. The choice of geothermal power is deliberate: Kenya’s electricity grid is 90%+ renewable through its geothermal and hydro assets, giving the facility a sustainability profile that hyperscalers can apply toward their net-zero commitments while serving a market that is commercially meaningful.
Oracle is separately setting up public cloud regions in Kenya, Morocco, and additional African markets. Equinix has committed to investing $390 million over five years to expand data center operations across Nigeria, South Africa, and East Africa. The investment volumes are not speculative projections — they are committed capex with named timelines and announced partnerships.
What Countries Must Do to Qualify
Hyperscaler investment does not follow markets passively — it follows prerequisites. The hyperscaler test analysis identifies the structural conditions that converted Kenya and Malaysia from candidates to destinations: reliable power, policy stability, infrastructure readiness, proven demand, inter-ministerial coordination, and local talent.
Power is the single hardest constraint. The IFC’s Global Sector Lead Obinna Isiadinso identifies reliable electricity as “the single most important constraint affecting data-center expansion in many emerging markets.” Kenya qualifies because its grid is stable and predominantly renewable. Nigeria has the demand but its grid unreliability has historically constrained hyperscaler confidence — Equinix’s investment there is partly predicated on self-contained power solutions.
Policy stability is the second filter. Malaysia attracted Microsoft’s first cloud region announcement in 2025 — a $2.2 billion commitment with three data centers planned for greater Kuala Lumpur — through a combination of tax incentives, regulatory clarity, and a 2026 budget offering foreign firms a tax holiday through 2047. India deployed over a dozen cloud regions through similar instruments. The lesson from both markets is that policy certainty matters more than market size: hyperscalers are making 15-20 year infrastructure bets and cannot absorb regulatory uncertainty on that timescale.
The third prerequisite is proven demand expressed through public-sector cloud adoption. Countries that have digitized government services, adopted cloud-first procurement policies, and published digital economy roadmaps provide the anchor-tenant signal that makes private investment commercially viable. South Africa, Kenya, and Nigeria each had this signal before their hyperscaler commitments materialized.
Advertisement
What Enterprise Leaders Should Do About It
The infrastructure buildout has practical implications for enterprise IT leaders — both in Africa and globally — who need to make cloud architecture and vendor decisions with a multi-year horizon.
1. Remap Your Africa Latency Assumptions for 2026 and Beyond
Decisions made 18 months ago about where to deploy Africa-facing workloads were based on an infrastructure map that no longer reflects reality. The Microsoft Kenya launch, the Oracle regional expansions, and the Equinix buildout across East and West Africa mean that latency-sensitive applications that previously required South African hosting may have better options by late 2026. Enterprise architects should re-evaluate workload placement for Africa-region deployments at 12-month intervals, not the traditional 3-5 year review cadence, during this infrastructure expansion phase.
2. Use Public-Sector Cloud Adoption as a Lead Indicator of Where Infrastructure Is Coming
The hyperscaler deployment pattern is consistent across India, Malaysia, Kenya, and Nigeria: government cloud adoption precedes hyperscaler investment by 18–36 months. Countries where digital services are advancing government-to-citizen interactions and where cloud-first procurement has been formalized are the next wave of hyperscaler targets. Enterprise IT teams with Africa-wide operations should track public-sector cloud adoption in Morocco, Egypt, Ethiopia, and the Francophone West Africa cluster as leading indicators of where infrastructure will land next.
3. Build Data Residency and Sovereignty Requirements Into Architecture Now
The hyperscaler test framework notes that countries seeking cloud regions increasingly require data residency guarantees. Nigeria’s Data Protection Act (2023), Kenya’s Data Protection Act (2019), and South Africa’s POPIA (active from 2021) each have data localization implications for enterprise deployments. As hyperscaler regions proliferate across Africa, the legal and technical requirements for data residency will become more complex, not simpler. Enterprise architects who bake sovereignty and residency compliance into their cloud architecture now — rather than retrofitting it when regulators demand it — will avoid the expensive remediation cycle that is playing out in European markets post-GDPR.
4. Evaluate Inference Infrastructure Before Training Infrastructure
The hyperscaler test analysis makes a clear sequencing recommendation for Africa: countries should focus on inference infrastructure and distributed compute layers serving regional users before attempting massive AI training campuses. The practical implication for enterprise IT leaders is equivalent: deploying AI model inference locally in Africa is increasingly viable and economically justified given latency and data sovereignty requirements. Deploying training workloads in Africa is not yet cost-competitive with established hyperscaler clusters in Europe or Asia. Separate the two decisions accordingly.
What Comes Next
The $600 billion in global hyperscaler capex projected for 2026 — a 36% increase over 2025 — is being allocated primarily to the US, Europe, and major Asia-Pacific markets. Africa’s share is real but proportionally small. What is different in 2026 is that Africa is no longer competing purely on market-size arguments; it is competing on specific infrastructure advantages — Kenya’s geothermal power, Nigeria’s 220 million consumers, Morocco’s cable connectivity to Europe — that are not replicable elsewhere.
The eight data centers launched in 2026, the 24 nations with carrier-neutral facilities, and the Rwanda-Zimbabwe expansion timeline represent the second act of Africa’s cloud buildout. The first act established South Africa as the continent’s cloud hub. The second act is establishing a multi-node infrastructure footprint that serves African enterprise demand from within the continent. For the countries that have done the policy and power work — and for the enterprises that have mapped their Africa latency and sovereignty requirements — that second act is already underway.
Frequently Asked Questions
Which hyperscalers are currently investing in Africa outside South Africa?
Microsoft committed $1 billion for a geothermal-powered 100 MW data center in Kenya (through G42 partnership, targeted for 2026). Oracle is establishing cloud regions in Kenya, Morocco, and other African markets. Equinix committed $390 million over five years for Nigeria, South Africa, and East Africa. Eight new data centers launched across Africa in 2026 alone, expanding the carrier-neutral hosting footprint to 24 countries. AWS currently operates only the Cape Town region but has expansion plans underway.
What does a country need to attract a hyperscaler cloud region?
The six measurable prerequisites are: reliable power supply (the single hardest constraint in emerging markets), policy and regulatory stability over 15-20 year investment horizons, pre-cleared infrastructure sites with grid transmission capacity, proven public-sector cloud demand acting as anchor tenant, inter-ministerial coordination across power, land, tax, and telecoms, and local talent pools for operations and compliance. Kenya qualified on all six; Malaysia followed with a $2.2 billion Microsoft commitment after offering a tax holiday through 2047 and demonstrating similar prerequisites.
How is Africa’s cloud buildout different in 2026 compared to 2020?
In 2020, Africa’s enterprise cloud infrastructure was effectively a single node: the AWS Cape Town region. In 2026, there are 24 nations with carrier-neutral data centers, eight new facilities launched this year, and billion-dollar commitments from Microsoft, Equinix, and Oracle targeting Kenya, Nigeria, Morocco, and East Africa. The buildout has shifted from South Africa as a single hub to a multi-node continental infrastructure serving Africa-regional workloads with materially lower latency and stronger data sovereignty compliance posture.
Sources & Further Reading
- Powering Africa’s Digital Future: Cloud Data Centres and Network Resilience — Console Connect
- The Hyperscaler Test: What Countries Must Do to Attract Global Cloud Regions — Africa Hyperscalers
- Hyperscalers in 2026: What’s Next for the World’s Largest Data Center Operators — Data Center Knowledge
- Cloud Platforms in Africa for Developers — Launchverse













