Microsoft's 74-Country Outage Exposed Africa's Banks Have No Plan B
Microsoft's 74-country outage became 2026's largest disruption, hitting Equity, Absa and Wema Bank — and African banks' only backup was WhatsApp.
By Amara Osei
4 min read

What's News
- The Microsoft outage began around 4 PM GMT on a Thursday and affected users in 74 countries.
- Downdetector recorded it as the largest outage across all monitored services in 2026.
- At least six Kenyan lenders, including Equity and Standard Chartered, sent customer notices.
- More than 15 hours passed between Microsoft acknowledging the fault and some lenders notifying customers.
- Copilot also went down during the same outage.
A Microsoft outage that hit users in 74 countries became the largest service disruption of 2026 by Downdetector's count, and it left African banks improvising their way through a Friday morning with nothing but phone numbers and WhatsApp messages.
The outage began around 4 PM GMT on a Thursday. Microsoft blamed settings applied to servers after an update: a process that sets up secure connections became overwhelmed, servers stopped handling requests, and the company began rolling the change back.
For TechCabal senior reporter Kenn Abuya, the failure landed at a personal level. He was at Avenue Hospital in Nairobi on Friday morning when the screens went blank. Nobody with insurance was going anywhere — the hospital needed the insurer's approval, and the insurer's approval ran through Microsoft systems. By lunch, his phone filled with bank texts saying the same thing: our email is down, please call us.
Which banks were hit?
The timing was brutal for African businesses. Outage reports climbed between 8 AM and 11 AM local time on Friday, just as offices across the continent opened. The list of affected institutions stacked up quickly:
- Kenya: at least six lenders sent notices, including Equity, Standard Chartered, Stanbic, UBA, I&M, Loop, and NCBA's digital bank
- Nigeria: Wema Bank and Stanbic IBTC told customers their email services were having problems
- South Africa: Absa, Eskom, FlySafair and the SABC were hit, and the South African Revenue Service warned the outage was affecting its services
Microsoft first acknowledged the problem on Thursday evening. I&M posted its notice on Friday; Loop texted customers by midday. More than 15 hours passed between Microsoft admitting the fault and some lenders telling customers where to go instead.
What was the backup plan?
Nearly every notice offered the same escape route: call us, message us on WhatsApp, or ring your relationship manager. Wema advised customers with urgent requests to use its phone lines, WhatsApp and Instagram — and two of those three channels belong to Meta.
That's the structural problem, Abuya argues. Banks dealt with one American platform failing by sending customers to another. "Meta is no backup for Microsoft," he writes. "Both are American giants with their own outage histories, and a bank leaning on one when the other falls over has spread its risk between two companies, not away from them."
The banks' notices stressed that payments and mobile banking were unaffected. No Kenyan lender reported an interruption to either service. But much of the routine work at banks now runs through email — loan applications, bank statements, customer complaints. Payments can keep moving while everything else grinds to a halt.
Why didn't banks build a fallback?
Running without a real fallback might still be the rational choice, the report concedes. A second email system that is tested and staffed costs money every month and pays off perhaps one day a year. A shared outage also protects everyone's reputation: when banks in three countries go down together, customers don't move their accounts and no executive loses a job.
The result is governance theatre. "The continuity plan promises resilience, the regulator files it, the auditor signs it off, and on the day, it fails; the plan turns out to be a phone number," Abuya writes.
Anyone hoping a local data centre would have saved the day should look at Kenya's planned Microsoft–G42 facility, which was supposed to host an East African Azure region and stalled after the government refused the payment guarantees the companies wanted. Even if it were running, servers in Olkaria installing the same bad update would have failed just as hard. Where the servers sit was never the problem; depending on one company is.
The next outage will land on offices that look very different. Microsoft wants Copilot and its AI agents inside the same corporate accounts that went dark, doing work like approving insurance claims and answering customers. Copilot also went down on Friday. If an insurer has moved its claims desk onto a Microsoft 365 agent by 2028, a similar outage means claims stop being processed — and the staff who once did that work by hand may have been let go.
Africa got through Friday because of people: receptionists, call-centre agents and bankers answering WhatsApp messages on their own phones. Automation is built to remove exactly that spare human capacity, and boards and central banks have just been handed a free fire drill. When the agents have taken over the desks and the next outage arrives, the report's closing question is who will be left to pick up.
Source: techcabal.com
More from Amara Osei
Show full bio
Senior reporter covering companies, deals and strategy in entrepreneurship, startups and venture funding at Business Bearings.
2,136 articles