On a Thursday afternoon in mid-November, the everyday machinery of email and workplace messaging faltered for millions of people at once, as Microsoft's Outlook and Teams, along with its Copilot assistant, went dark across much of the world. The failure reached beyond Microsoft itself, cutting off the connectors that link OpenAI and Anthropic tools to corporate accounts, and reminding us how deeply our work now rests on a few shared pipes. Microsoft has said it is working to restore service, but its efforts have not yet held, and the full account of why a system built to carry so much could be
Microsoft Races to Restore Outlook and Teams as Global Outage Hits Dozens of Countries
Restarted servers stopped responding again once traffic returned
On a Thursday afternoon in mid-November, the everyday machinery of email and workplace messaging faltered for millions of people at once, as Microsoft's Outlook and Teams, along with its Copilot assistant, went dark across much of the world. The failure reached beyond Microsoft itself, cutting off the connectors that link OpenAI and Anthropic tools to corporate accounts, and reminding us how deeply our work now rests on a few shared pipes. Microsoft has said it is working to restore service, but its efforts have not yet held, and the full account of why a system built to carry so much could be brought down by its own recovery remains to be written.
Shortly after 1 p.m. Eastern on Thursday, Outlook stopped working for a vast number of people. Downdetector, which tracks reports submitted by users, recorded close to 130,000 complaints about Outlook in the United States within moments. Microsoft soon confirmed it was working to fix a global outage across its Microsoft 365 apps, and within about an hour reports had come in from 74 countries and territories.
For different people the failure looked different. Some could not send or receive email, while others found the program unreachable altogether. Teams, the workplace messaging app, went down as well, along with other services including the Copilot AI assistant.
The damage did not stop at Microsoft's own doors. Connectors that let OpenAI and Anthropic tools work with Microsoft 365 accounts also stopped functioning as expected. For organizations that tie their AI tools to email and workplace accounts, one failure rippled into another.
Microsoft's first statement, posted on X, said it knew users were struggling to reach Exchange Online mailboxes in the Americas and blamed elevated load on part of its infrastructure. The recovery then faltered. Some restarted servers stopped responding once traffic returned, and at 0254 GMT on Friday the company admitted its latest changes had not worked as expected.
That update came about nine hours after the first reports, and the outage was still not resolved. The available reporting does not say how many users were still affected, and the explanation of elevated load remains preliminary. What comes next is Microsoft's full root-cause analysis and whether service holds steady as traffic returns.
Shortly after 1 p.m. Eastern on Thursday, a little after 1800 GMT, Outlook stopped working for a very large number of people. Downdetector, the service that tracks user-submitted outage reports, logged nearly 130,000 reports of Outlook problems in the United States alone within moments of that mark.
Microsoft said on Thursday that it was working to fix a global outage affecting its Microsoft 365 apps, Outlook and Teams among them. Within about an hour, Downdetector had received reports from 74 countries and territories. The scale of the reporting suggests a problem well beyond one region or one customer group.
What users experienced varied. Some said they had trouble sending and receiving email. Others said the program could not be reached at all. Outages of the Teams workplace messaging app were also reported, along with problems in other Microsoft 365 services, including the AI assistant Copilot.
The disruption did not stay inside Microsoft's own products. Microsoft 365 connectors used by OpenAI and Anthropic were also affected, and both companies said their tools were not working as expected. For organizations that link their AI tools to their email and workplace accounts, the failure of one service had consequences for another.
Microsoft's first public acknowledgment came on X. The company said it was aware that users were experiencing issues accessing Exchange Online mailboxes through multiple connection methods in the Americas. It said part of its infrastructure was experiencing elevated load, which it described as likely contributing to connectivity errors. It added that it was aggressively pursuing all pathways to reduce the impact.
The recovery did not go smoothly. Microsoft said some servers that had been restarted stopped responding again once traffic returned to them. In other words, systems that came back up could not hold under the demand placed on them. In an update at 0254 GMT on Friday, the company said its latest changes had not worked as expected and that it was looking for other ways to restore service.
That update came roughly nine hours after the first wave of reports, and it left the outage unresolved. The source reporting does not say how many users remained affected at that point, and Microsoft's description of elevated load is so far a preliminary explanation rather than a finished account of the cause.
What to watch next is Microsoft's full root-cause findings, and whether service returns without further relapses as traffic comes back.
Notable Quotes
We are aggressively pursuing all pathways to reduce impact.— Microsoft, on X
Want the full story?
Read the original at Deutsche Welle ↗