For more than a full day, the tools most of the working world runs on simply stopped working. Microsoft's Outlook, Teams and Copilot went dark for millions of people on Thursday afternoon, and the outage dragged on for more than 24 hours before the company said its services were fully restored early on Saturday morning.
More than 430 million people use Microsoft 365, the bundle of workplace apps caught up in the failure, and the trouble was reported in at least 74 countries. Downdetector, the site that tracks outage reports, recorded a peak of about 130,000 reports about Outlook and roughly 50,000 about Microsoft 365 on Thursday afternoon. The real number of people affected was almost certainly far higher: Downdetector only counts the people who bother to report a problem.
A Network Upgrade Gone Wrong
Microsoft said the failure was a technical issue introduced during a network capacity upgrade, which surfaced after a recent network change. In the company's words, the incident was introduced by an issue during an earlier network capacity upgrade that surfaced following a recent network change — a self-inflicted problem, not an outside attack.
That distinction mattered, because the timing invited suspicion. The outage began on Thursday, the same week as a nationwide FAA ground stop, and speculation spread online that Microsoft had been hacked, or that an artificial intelligence agent had somehow broken the company's own systems. Microsoft addressed that directly on Friday morning, saying there was no indication the outage was the result of malicious or cyber activity, and later adding that it was a technical issue not related to cyber or AI agents.
The company also said it would publish a full post-incident report once its investigation is complete, and that the investigation is still ongoing.
Hospitals, Courts And Offices Felt It First
The outage hit the apps people use to do their jobs: Outlook email, Teams meetings, Copilot, OneDrive, OneNote and To Do. Businesses, hospitals, courts and government offices in more than 70 countries reported disruption, with workers locked out of their inboxes and meetings in the middle of the working week. Microsoft extended its email timeout window to 48 hours so delayed messages would still get through instead of bouncing back.
Teams, the meeting tool that has become the default conference room for millions of workers, was among the services affected, along with Copilot, the AI assistant Microsoft has spent the last two years putting at the center of its workplace software. For companies that run their day through those tools, the outage did not just remove email: it removed the meeting room, the shared files and the assistant, all at once.
For people trying to work through it, the recovery came in waves. Microsoft reported only partial recovery in some parts of the world by the early hours of Friday, then broader improvements across North America and Europe by early Friday afternoon. By 3 p.m. ET on Friday, the company's service health page showed its main services, including 365 and Outlook, as operational, and reports on Downdetector had fallen to about 1,600 for Outlook, down from the 130,000 peak.
Even then, the company warned that some users could still see problems, with email delivery lagging behind service access and some features taking longer to stabilize. Early on Saturday morning, Microsoft said the work to fix the underlying problem was complete and customers using its main workplace tools should no longer be affected.
A $7 Billion Bill For Lost Work
The outage could cost about $7 billion in lost productivity, according to an estimate by Anderson Economic Group, a Michigan-based research firm. The figure captures a simple reality: when email and meetings stop for a day, the work does not just pause, it piles up, and the hours spent catching up are hours not spent on anything else.
Microsoft apologized on Friday afternoon in a post on X, writing that it was deeply sorry for the disruption the incident had caused its customers and partners, and that it was seeing broad recovery and stability across its services. The company said it restored services gradually, validating stability at each stage, because moving faster risked breaking things again.
What To Do If Your Email Is Still Missing
If you use Outlook and something sent during the outage still has not arrived, the most likely explanation is the backlog: Microsoft said queues of emails waiting to be delivered were significantly reduced and approaching normal levels by Friday evening, and the 48-hour timeout extension was designed to protect delayed messages. Check the junk folder, then the web version of Outlook, and give it a little more time before assuming a message is lost for good.
Outages of this size are rare for Microsoft, but they are not new for the industry. Cloud services fail, and when they do, the failure is felt everywhere at once, because so many businesses now depend on the same handful of providers. That is why the post-incident report matters beyond Microsoft: every company that lived through Thursday will want to know exactly which safeguard failed, and what replaces it.
The bigger question is what Microsoft's report will say when it lands. The company has promised a full account of what went wrong during the network capacity upgrade, and of what it will change so a routine upgrade cannot take down the working world's inbox again.
