Microsoft outage hit 74 countries, and the company says it wasn't AI

Microsoft says a botched network upgrade, not hackers and not artificial intelligence, knocked out its services across 74 countries and territories — disrupting hospitals, airlines, visa offices, schools and banks before a "broad recovery" took hold on Friday.

The company apologized directly on X: "We are deeply sorry for the disruption this incident has caused our customers and partners." And it moved quickly to kill the speculation that always follows an outage of this scale: "We are seeing broad recovery and stability across our services. This was a technical issue not related to cyber or AI / agents."

What actually broke?

The timeline matters here. On Thursday, when the failures began, Microsoft attributed the disruption to "elevated load" on its infrastructure. By Friday, the company had a more precise diagnosis.

The "incident was introduced by an issue during an earlier network capacity upgrade that surfaced following a recent network change," Microsoft said in its updated statement. In other words, a latent fault sat dormant inside Microsoft's network until a routine change triggered it.

"We have addressed the underlying issue and continue to restore service to remaining customers and regions," the company said.

The blast radius widened over the course of the incident. Microsoft initially said it was investigating outage reports "in the Americas." It later confirmed the issue apparently affected "Microsoft 365 service connectivity for users in any region." Microsoft 365 is the company's cloud-based subscription service, and it bundles the applications much of the working world runs on — Outlook and Teams chief among them.

The outage tracker Downdetector recorded disruptions for users across the Americas and parts of Europe, Asia and Africa. By 0900 GMT on Friday, problem reports had begun to decline.

Where did the outage hurt most?

The consequences landed hardest on public services, and the specifics pile up fast:

• United Kingdom: National Health Service staff receiving emergency calls were "unable to refer patient details in the normal way," creating a massive backlog, said Rachel Harrison, public services head at the GMB union. The union described a "serious impact" on the country's health care service.
• New Zealand: Immigration New Zealand said the disruption forced a halt in the processing of all visa applications.
• India: Airlines IndiGo and Akasa Air both said on X that their email support services had been affected and that they may not be able to receive customers' emails.
• Japan: Economy, trade and industry minister Ryosei Akazawa warned of a "significant impact" because Microsoft products are used by a "vast number of people."
• Global business: The disruption delayed corporate earnings reports and left schools struggling to communicate, according to the company's own account of the incident's reach.

A grounded airline passenger can wait. A patient referral backlog and a frozen visa system are different categories of harm, and Thursday's outage delivered both.

What did the people in the middle of it say?

In London's financial district, several people interviewed by AFP said the outage exposed something uncomfortable: how much their daily work now depends on Microsoft's cloud — and, increasingly, on Copilot, the company's AI assistant.

"I think one of the biggest things is Copilot not working," said Yashmit Sharma, 26, an auditor. "I think this was the main bummer. And Outlook as well, of course."

Sheera, 40, a risk manager at a mid-sized bank who declined to give her last name, said businesses need to be better prepared for unexpected technology failures. "I think we need to be a little bit more careful," she said. "Anything can happen: you wake up and Outlook is not working and it's the main communication we use for our global team."

That quote cuts to the operational reality of modern work. When a single vendor's email and collaboration stack goes down, a global team's primary channel goes down with it — and most mid-sized organizations have no second channel standing by.

Why does a single Microsoft fault matter to whole countries?

The incident is already producing policy aftershocks. The Australian Computer Society called for a national review of sovereign IT risk, arguing that concentration on foreign cloud platforms creates national exposure.

ACS CEO Prins Ralston pointed to small businesses as the most vulnerable layer. When Outlook fails, he said, they may simply have no workaround.

"We need to know the systems that keep this country running won't simply fall over without notice," Ralston said, according to the ACS publication Information Age.

That framing turns a one-day outage into a structural question. Hospitals, immigration systems, airlines and government ministries all sat on the same Microsoft infrastructure, and they all went dark together. Microsoft's own statement — emphasizing that no "cyber or AI / agents" activity was involved — underlines the point: the failure was mundane, a capacity upgrade gone wrong, and it still cascaded across 74 countries and territories.

What comes next?

Microsoft says it has fixed the underlying issue and is restoring service to remaining customers and regions. Downdetector's declining report curve suggests the worst has passed. But the scrutiny will outlast the outage itself: with regulators and industry bodies like the ACS now asking how sovereign economies can depend on a single vendor's uptime, this week's technical fault has handed cloud-reliance skeptics their most concrete case yet.