Microsoft Outlook, Teams and Microsoft 365 users reported problems in 74 countries from about 20:00 SAST on Thursday, and repairs were still being rolled out on Friday morning.
EWN reported that Downdetector, an outage tracker, logged problems in 74 countries and territories within an hour, including parts of Africa. In the United States, about 130,000 users reported Outlook problems shortly after that time. By 01:00 SAST on Friday, about 37,000 were still reporting them, EWN reported.
Microsoft said “elevated load” on part of its infrastructure was “likely contributing” to the connection errors. SABC News reported that several servers unexpectedly went offline, which put more pressure on the servers still running, and that Microsoft’s automated health checks appeared to be adding to the problem.
The Guardian reported that Microsoft limited traffic to the affected servers, and that restored servers failed again once traffic returned. At 05:30 SAST on Friday, Microsoft said Exchange, the service behind Outlook email, was the most affected, with Teams chats and meetings less so. It was also assessing whether to revert recent configuration updates, the Guardian reported.
A Microsoft spokesperson said the company “had no indication of a cyber-attack”, the Guardian reported. Charles Lamanna, Microsoft’s executive vice-president for Copilot, wrote on X that there was “no indication this is the result of malicious or cyber activity”. According to Al Jazeera, Australia’s home affairs minister, Tony Burke, said Microsoft believes it is dealing with something “in some ways similar” to the CrowdStrike failure, “where there is an update that has gone awry”.
Al Jazeera reported that problems also appeared in Asia and Australia on Friday morning, and that Microsoft said its recovery was not going as well as expected. At 07:43 SAST, SABC News reported that an initial fix had restored some servers and was being rolled out more widely, and that the earlier Teams disruption was resolved.
The Guardian said the failure also reached the Microsoft connectors used by OpenAI’s and Anthropic’s AI services. It added that Microsoft reported an outage on its Azure cloud platform last month.
Mark Gregory of RMIT University said in the Guardian that when some servers have configuration problems, every customer is pushed onto a smaller pool, which overloads it. He said firms that depend on cloud services should keep an alternative, such as a second email account.
Microsoft has said it will share more information when it has it. TechCentral reported on Friday that the outage began at about 20:00 SAST on Thursday, and Outlook, Teams and Copilot were still down on Friday. Microsoft said it was seeing signs of stabilisation as fixes were applied more widely.
The City of Cape Town said there was “no confirmed timeline for the full restoration of services”. Email was unreliable for city staff, Teams dropped in and out and some had trouble reaching OneDrive files, TechCentral reported. SARS issued a warning that its taxpayers might find Outlook, Teams and other Microsoft 365 tools disrupted. TechCentral’s report does not say whether eFiling was affected. The University of Cape Town’s ICT department said Microsoft had told it of connection problems on several Microsoft 365 services.
Microsoft said in a service health notice that features that depend on Exchange Online and email were affected, while those that do not rely on Exchange Online still worked. Downdetector South Africa logged 874 Microsoft 365 complaints at 07:53 and 744 Outlook complaints at 07:38 on Friday, which count user complaints, not affected accounts. None of the reports saved here gives a time for full service to return.