Last week’s Microsoft outage, which cut users in Africa and around the world off from email for almost 24 hours, began with a software fix rolled out in July, according to a preliminary report Microsoft sent to customers on Sunday.
The Microsoft outage findings matter for African banks, insurers and public bodies that run daily operations on Microsoft 365. In Kenya, several large lenders moved customers onto WhatsApp and phone lines; in South Africa, the revenue service and the City of Cape Town warned of disruption.
What caused the Microsoft outage?
The Microsoft outage started with a change written for servers in a networking gateway that sits in front of many Microsoft 365 services, including Exchange Online, the email system behind Outlook, TechCentral reported, citing the report. Under certain loads, it caused tasks inside a Windows process on those servers to wait on one another until the machines ran out of processing power and memory.
The change had passed through Microsoft’s staged deployment from July. “For more than three months, the change was deployed worldwide without impact,” the report said. Microsoft said its load testing had missed the fault because the problem is worse on newer machines with more processor cores, while its tests “tended to focus on weaker hardware with fewer cores”. The technology site Neowin also reported that the failure traced back to July.
How did the Microsoft outage spread so far?
On Thursday, 12 November, Microsoft’s systems automatically moved Teams traffic onto the gateway that also carries Exchange Online, at a time when nearly 14% of that gateway’s capacity was already out of action because of an unrelated problem. The extra load triggered the fault during peak US hours.
As servers stopped responding, Microsoft’s traffic management sent users to healthy servers, which then failed in turn. “This caused a cascading failure from server to server as the same amount of load was spread across fewer servers,” Microsoft said. At the low point, more than 80% of the gateway’s servers were unresponsive. Barriers meant to contain failures to one region did not hold, because of what Microsoft called a separate issue that the preliminary report does not explain.
Because the faulty code was months old, its usual rollback tools could not quickly undo it. Microsoft began reverting the change about 15 hours after the first alert, prioritising regions entering business hours. Webmail was restored worldwide by 4.17pm South African time on Friday, 13 November.

How did the Microsoft outage hit African businesses?
In South Africa, the South African Revenue Service and the City of Cape Town warned users of disruption, as DNE Africa reported on Friday. In Kenya, Equity Bank and UBA Kenya asked customers to call or use WhatsApp, Stanbic Bank Kenya pointed them to its call centre and Standard Chartered told customers its email was unavailable, TechCabal reported. Old Mutual Kenya and APA Insurance also lost email. None of the banks’ notices mentioned problems with payments.
Downdetector, owned by Ookla, said the Microsoft outage was the largest it had recorded in 12 months, with more than 7mn reports and more than 90 countries or territories logging at least 1,000 each, according to Ookla’s analysis. Reports stayed high for about 20 hours, twice as long as any other large outage in that period.

What happens next?
Since the Microsoft outage, the company said it had added monitoring of how traffic is redirected, limited how load is redistributed and added capacity. It plans to publish a final report with a full list of repairs by 20 November.
Customers may claim service credits under Microsoft’s service level agreement, which pays 25% to 100% of the affected service’s monthly fee depending on downtime, but claims must be filed by 31 December and credits do not cover lost revenue. Kenya has few local alternatives: a planned $1bn Microsoft and G42 data centre meant to host an East African cloud region has stalled over payment guarantees, TechCabal reported, although officials say it remains on track. South Africa is meanwhile courting US technology investment.
FAQ
What caused the Microsoft outage in November 2026?
Microsoft’s preliminary report traces it to a change made in July to servers in a networking gateway in front of Microsoft 365, which failed under heavy load and spread through its failover systems.
How long did the Microsoft outage last?
Almost 24 hours, from 6.20pm South African time on 12 November to 6.10pm on 13 November, according to Microsoft’s timeline as reported by TechCentral.
Was it a cyberattack?
No. Microsoft said it found no evidence of a cyberattack.
Which African organisations were affected?
They included South Africa’s revenue service and the City of Cape Town, and in Kenya Equity Bank, Standard Chartered, Stanbic, UBA Kenya and Old Mutual Kenya, which lost email or redirected customers.

