One Company's Network Upgrade Froze Governments Worldwide

Immigration New Zealand paused processing of all visa applications. That sentence is the whole story. A single company's botched network upgrade, and a state agency that decides who may cross a border simply stopped, because the email wouldn't send.

Microsoft apologized Friday for a global outage that knocked out Outlook, Teams, the Copilot AI service and other Microsoft 365 services for millions of people, companies and governments. The cause, according to the company: "an issue during an earlier network capacity upgrade that surfaced following a recent network change." It wasn't a cyberattack, Microsoft said, and it wasn't artificial intelligence agents. On Thursday it had told the public: "As of now, we are not aware of any malicious activity that is causing this incident."

The Apparatus Runs on One Switch

The outage began Thursday, November 12, and continued into Friday, November 13. In that time the damage spread across continents. In Australia, Service NSW, state and territory departments and Queensland Health were reportedly affected, along with telehealth, schools and business communications. Japan's economy, trade and industry minister, Ryosei Akazawa, said the outage would have a "significant impact." OpenAI and Anthropic reported linked problems. Indian carriers IndiGo and Akasa Air had email support disruptions, though Air India said flight operations were unaffected.

More than 430 million people use Microsoft 365 apps, according to reports. Downdetector recorded about 140,000 Outlook problem reports from 74 countries Thursday evening. Analysts told The National that a continuing outage would have a "cascading" effect on businesses.

Read that list again. Health departments, schools, visa offices, ministries and airlines: institutions that claim sovereignty over borders, bodies and budgets, all routed through the mail servers of one private corporation. Every one of them chose the dependency. None of them could work around it when it failed.

Apologies From the Top

Microsoft's account of itself shifted as the hours passed. It first said the issue was limited to the Americas, then warned it could affect users "in any region." It blamed "elevated load" on part of its infrastructure, then connectivity errors in part of its Exchange Online infrastructure, then a process responsible for establishing secure connections on affected servers that became overwhelmed. It paused deployments, moved traffic away from affected infrastructure and restarted servers. On Thursday it reported only "limited recovery."

By Friday the tone was contrition. "We are deeply sorry for the disruption this incident has caused our customers and partners," the company said, adding: "We know the trust they place in our services and take that responsibility extremely seriously." A full post-incident report is promised once its review is complete. Note who gets to review: the company that caused the failure.

Who Waited

In the UAE, users began receiving emails again from about 5.20pm local time Friday on desktop and mobile Outlook. Microsoft had not confirmed full restoration. Downdetector.ae showed reports first spiking near 300 around 10.30pm Thursday, then a larger wave from about 6am Friday peaking above 550 just after 9am, and still above 200 into the afternoon. Of reported problems, 42 per cent involved the Outlook app, 29 per cent receiving messages and 18 per cent logging in.

A separate Dynamics 365 Contact Center disruption (CR1550885) lasted nearly eight hours, with access restored at 00.57 GMT Friday.

Emirates posted an advisory on its website warning that booking confirmation emails may take longer than usual to arrive. It reassured customers their reservations remain secure, said no additional action is needed, and apologised for any inconvenience. One corporation apologizing for another's failure, both asking the public to wait patiently.

Microsoft said it restored services gradually, "validating stability at each stage," and acknowledged that some users might still face "intermittent issues" or delayed email as "mitigation progresses across the remaining affected infrastructure." Nobody who lost a day of work to this gets a vote on the architecture. Nobody does.