Microsoft Outlook outage spreads worldwide as servers unexpectedly drop offline

Microsoft's Outlook email service went down for large numbers of business and consumer users on Thursday. The company traced the trouble to "a large number of servers being unexpectedly unavailable," which shifted extra load onto the machines that were still running.

The damage reached well beyond email. Microsoft said Teams, SharePoint, Copilot and its Defender XDR security product were also affected, and consumers were cut off from OneDrive storage and Office for the web. As of Thursday afternoon, the company had not said what knocked the servers out or when service would be fully back.

Complaints began piling up on Downdetector in the early afternoon Eastern time. Within about an hour, Outlook reports on the site had passed 100,000. Microsoft first described the incident as a problem reaching Exchange Online mailboxes "through multiple connection methods in the Americas." It later broadened the alert to users in any region.

Microsoft is tracking the incident as EX1550543 in its Microsoft 365 admin center, where IT administrators normally go for details. TechCrunch said the admin center was unreachable when it published. Microsoft's public service status page was also slow to load, or failed to load at all, for much of the afternoon. The company turned to its Microsoft 365 Status account on X to post updates.

That is a real weak spot. A status page is most valuable during exactly this kind of event. When it struggles alongside the services it reports on, administrators are left piecing together the situation from social media posts and crowd-sourced outage trackers. The X account did the job on Thursday, but it is a fallback, not a plan.

What Microsoft says went wrong

Microsoft's explanation changed over the afternoon. Its early updates said part of its infrastructure was under elevated load in multiple regions, and that the load was likely behind the connection errors users were seeing. Engineers responded by spreading connection traffic across Microsoft's worldwide infrastructure, moving more traffic off the affected systems and restarting servers to bring them back into service.

At that point the company said it was still trying to work out whether the surge was a real jump in requests or a change in how traffic was being routed. It later ruled out an overall increase in traffic. Instead, it said, servers had become unavailable in large numbers, and the rest of the system was absorbing the work they would normally handle. Microsoft said it was investigating why the servers dropped out while working to return them to service safely.

Microsoft's own account describes a cascading failure. When enough servers fall over, the remaining ones take on their load until they start to strain too. Shuffling traffic and restarting machines can relieve that pressure. It does not address whatever took the servers down, and until Microsoft finds that cause, a relapse is a reasonable worry.

Microsoft also said it had paused all deployments and configuration changes as a precaution. Engineers are reviewing recent changes to see whether any of them line up with the failure. "We're aggressively pursuing all pathways to reduce impact," the company said.

The company has not said a change caused the outage, and that question is still open. Still, a freeze on every deployment is a costly step for a company of Microsoft's size, and taking it suggests a bad update is one of the leading possibilities. That would carry extra weight given recent history. In September, a Microsoft 365 outage lasted for days, and Microsoft blamed a "core authentication configuration" issue for keeping users out. If Thursday's failure also traces back to a change Microsoft made itself, customers will reasonably want to know what is different about how the company tests and rolls out updates.

How far the disruption spread

Downdetector also showed spikes in user reports for Azure, Microsoft Store, Microsoft 365 and Copilot. Cloudflare, Google and Gmail appeared on the site as well. Those reports come from users, and they can mislead: someone who can't reach one service often assumes others are failing too. TechRadar said it found no other sign that Google's services or Cloudflare were having similar problems, and that Amazon Web Services' status page showed no widespread outages.

The effects inside Microsoft's own ecosystem were real enough. TechRadar reported that some organizations, including universities, had lost access to their entire email systems. The outlet said OneDrive's website, which it could reach earlier in the afternoon, later stopped responding.

Microsoft's reach explains why one infrastructure failure was felt so widely. Many organizations buy email, chat, file storage, document editing and security tools from the same vendor, in the same bundle. That arrangement is convenient on a normal day. On a day like Thursday, email, meetings and shared files can all go down together, leaving workers with no obvious backup channel. Defender XDR's appearance on the list of affected services stands out, because security teams need their tools most during disruptions.

There were early signs of easing. By 2:25 p.m. ET, Downdetector showed a small drop in outage reports across Microsoft services, though complaints were still pouring onto social media. Microsoft has not said whether that dip reflects recovery or simply fewer people checking in.

Even reporters covering the outage felt it. TechCrunch said it would contact Microsoft for comment once its own email service was working again.

For now, the most useful clue about where Microsoft is looking is the deployment freeze. The company said it is reviewing recent changes for any link to the failure. What that review finds will matter more to Microsoft's business customers than how quickly inboxes come back this afternoon.