At 2:39 AM, offices worldwide came to a halt…MS365 suffers massive outage

At 2:39 AM Korean time on the 13th, Microsoft's cloud began to falter. By morning commute time, the aftereffects engulfed offices across the nation. With core MS365 services—Outlook (email), Teams (messaging and conferencing), and SharePoint (documents)—all going offline simultaneously, morning operations at domestic companies using these tools ground to a halt one after another.

Outage strikes at dawn, sweeps commute time

A major domestic conglomerate issued an emergency notice to all staff in the morning, while office workers complained that "morning operations were paralyzed as we had to re-contact dozens of business partners via personal email." Even after recovery progressed, aftershocks lingered. Cases were confirmed where emails sent at 5:30 AM arrived after 8:00 AM, with email delays continuing until 9:30 AM that morning. Around a maximum of 140,000 reports based on Outlook poured into outage tracking site DownDetector, and reports from Japan also surged after 4:00 AM.

Cause: "Server failure, not traffic surge"

According to Microsoft's announcement, the affected services span the full work stack—Exchange Online (Outlook), Teams, SharePoint, Copilot, and Defender. Microsoft explained the cause as "load resulting from multiple servers unexpectedly becoming inoperable, rather than increased traffic," and according to foreign media, recent configuration changes are being investigated as a potential cause (unconfirmed). Microsoft stated it has completed the transition to alternative infrastructure and is monitoring the situation, but as of 10:00 AM on the 13th, a full normalization announcement has not been made. The system is in a gradual recovery phase.

Fourth cloud crisis in 13 months

This outage reads more heavily due to the pattern. Within 13 months following the October last year Amazon Web Services (AWS) U.S. East Region outage, the late October Microsoft Azure Front Door outage (halting Starbucks and Alaska Airlines payments), and the November last year Cloudflare outage (disrupting ChatGPT and X), this marks the fourth major cloud incident. Whereas the previous three were failures of internet-based infrastructure, this time marks the difference that the office's everyday itself—email, conferencing, documents, and security—switched off all at once. It has been proven in just two hours from dawn that global corporate operations rest on the cloud of a single company.

As this publication noted in planning last week, stopping the everyday requires no missiles. A handful of inoperable servers can bring global offices to a standstill for half a day. What companies must examine is not whether to use cloud services or not, but the second path when they fail. Alternative contact channels, redundancy for critical documents, operational procedures during outages. This morning was the day that companies received that checklist.