Concerned analysts monitor a global cyber outage dashboard showing rising alerts across cloud, email, travel, and banking services.
Downdetector says last week was its busiest of the year. It logged 14 million outage reports worldwide as failures hit Microsoft, airlines, banks and other major services. Microsoft 365 was the biggest single contributor. The number says a lot about how people reacted to the outages. It says much less about how many people were affected or what broke.

Concerned analysts monitor a global cyber outage dashboard showing rising alerts across cloud, email, travel, and banking services. What the numbers say​

Fortune's reporting gives the clearest breakdown. Downdetector logged 6.4 million U.S. reports in the seven days ending Tuesday, four times a typical week, and 14 million worldwide, three times its usual level. Axios used a slightly different window. It reported nearly 6 million U.S. reports in the seven days ending Nov. 16, citing Ookla analyst Luke Kehoe. Those two U.S. figures are close but not identical, so treat them as separate snapshots. They are not a contradiction.

Three figures from the story are easy to mix up:

  • 14 million is the worldwide total for the whole week, across many services and incidents.
  • 7.8 million worldwide (about 3.9 million in the U.S.) is the Microsoft outage's share, according to Fortune's account of the November 12–13 incident.
  • 19,650 reports per minute was the Microsoft peak rate. It is a rate at one moment, not a count of affected users.

The Microsoft outage​

Microsoft's Outlook and Teams went down on Nov. 12 and 13, and Fortune says email was out in more than 70 countries for almost 24 hours. Fortune also says it was Downdetector's largest outage since the AWS incident in October 2025. Matching that scale does not mean the two had the same cause.

PCMag's running coverage, republished by Yahoo, adds detail on the peak. Kehoe said reports peaked at 19,650 a minute at 18:25 GMT, more than 1,500 times the normal level. Nearly 23 hours later they were still arriving at about 930 a minute. Outlook drew nearly two-thirds of reports. The problems people named were basic: emails not arriving, apps not working and sign-ins failing.

Microsoft's own explanation, as quoted by PCMag, was that the problem "was introduced by an issue during an earlier network capacity upgrade that surfaced following a recent network change." Microsoft repeatedly said it had no indication of malicious or cyber activity. The company also extended its email timeout window to 48 hours to protect delayed messages. Those statements cover this outage only.

One week, several unrelated failures​

Microsoft was not alone. Fortune lists these separate incidents:

  • FAA: a nationwide ground stop on Friday, which the agency tied to unstable legacy equipment.
  • Chase: a software bug on Monday, with about 90,000 U.S. reports. Fortune says that was roughly 165 times normal.
  • Apple: a faulty internal update on Tuesday, with about 81,000 reports.

Downdetector flagged 67 U.S. services that Tuesday. A typical day sees 16.

Mark Giles, director of Ookla Research, which owns Downdetector, called it largely coincidence: "There's nothing really linking the outages." Axios quoted Honeycomb's Liz Fong-Jones, who said the incidents have very little in common. She described the Microsoft failure as a cascading one. A software bug combined with increased traffic and reduced capacity, and the servers were overwhelmed as they tried to recover.

The patch-rush theory​

Washington has offered a different explanation. Fortune reports that Sen. Mark Warner told the Financial Times there was no sign of a cyberattack. He said the outages most likely came from organizations racing to patch software, with AI tools now finding flaws faster than companies can safely fix them. Fortune adds that neither Microsoft nor JPMorgan Chase tied its own outage to patching AI-found flaws. Axios likewise found no evidence of AI-assisted hacking so far.

Giles made another point. Downdetector can't identify root causes, so none of this shows up in its charts.

Why raw report counts can mislead​

Fortune notes that Downdetector counts complaints, not confirmed failures. Giles said the site can overstate an outage. When Chase payments failed, customers at other banks reported their own banks as down. Outage news also sends more people to the site, which inflates reports for unrelated services.

Fortune also puts the week in context. Downdetector sees one large outage, defined as more than 50,000 U.S. reports for one company in a day, a little more than once a week on average. Only Christmas week and one week in June had more than this week's four. Axios found the past 30 days had just three major outages, against a typical five. Giles said large outages have been falling since about 2022, once pandemic years are set aside.

What IT admins should take from it​

This is my analysis, not a finding from the sources.

  • Don't use Downdetector to judge your own tenant. It shows that users somewhere are complaining. It can't tell you whether your users are affected or whether your service is restored. Check the Microsoft 365 admin center's service health and your own telemetry.
  • Plan for email delays. Microsoft extended its email timeout window to 48 hours during this incident. Know what your own mail flow and any dependent systems do when delivery stalls.
  • Keep an out-of-band channel. When Outlook and Teams are both down, your incident communications can't depend on them.
  • Don't assume an attack. Every source cited here says the timing alone proves nothing. Treat "coordinated campaign" claims as speculation until an investigation shows a link.

Bottom line​

The 14 million figure is real, but it measures how much people complained. It does not measure how many were affected. The Microsoft outage was the largest single driver, and Microsoft attributes it to a network capacity upgrade and a later network change. The FAA, Chase and Apple incidents had their own stated causes. By Downdetector's own long-run data, this was a big cluster in a trend where large outages are becoming less common.

 

References

  1. Downdetector Logs 14 Million Outage Reports in Its Busiest Week - The Eastern Herald The Eastern Herald 2026-11-21T15:53:50+00:00
  2. Microsoft, FAA, Chase, and Apple outages drove 6.4 million U.S. Downdetector reports in one week | Fortune fortune.com
  3. ‘The whole internet screams at once’: Downdetector logged 6.4 million U.S. outage reports in a week as Microsoft, the FAA, Chase and Apple went down biztoc.com