Between Thursday and Tuesday, six large systems failed: Microsoft 365, the national air traffic system, Chase and the banks that depend on it, United Airlines, the National Security Agency's internal network, and Apple's iCloud. Investigators have found no sign of an attack in any of them. Each one broke while someone was changing it.
That is good news and bad news. Good news: it was not a hack!. Bad news: the systems your business depends on are complex enough that an ordinary upgrade can take them down for a day.
The week, in order
On Thursday, November 12, Microsoft 365 went down worldwide for about 24 hours. Email was down, along with calendars, Copilot, and every feature that depends on email. Microsoft traced it to a hidden Windows defect on the servers that handle connections into its service, set off by a routine shift in traffic.
On Friday, the Federal Aviation Administration stopped flights nationwide for hours, and thousands of flights were delayed or canceled. The cause was a new router installed during a network upgrade, which overwhelmed the older equipment around it.
On Monday, Chase customers had trouble with online banking, and the problem spread to several other major banks. Chase traced it to a software bug and had transactions moving normally by that night. The same day, United Airlines held departures for about an hour after its own system outage, which it has not yet explained.
On Tuesday, the National Security Agency's internal network for classified communications went down during scheduled maintenance, and the agency staggered staff arrivals while it recovered. The same day, Apple customers lost access to iCloud sign-in, the App Store, Apple Pay, and device activation, and later some lost iCloud Mail. Apple has not yet explained the cause.
What these failures have in common
The FAA failure is the cleanest example. A contractor replacing copper lines with fiber installed a new router in Arkansas. That router made roughly a thousand connection attempts every few seconds, more than the older equipment around it could handle. One device in one state grounded every flight in the country.
Microsoft's story follows the same shape at a larger scale. A defect sat unnoticed in its servers until a routine traffic shift exposed it. The servers that failed passed their load to the servers that remained, and those failed in turn. The NSA's network went down during scheduled maintenance. Chase traced its problem to a software bug.
The Secretary of Homeland Security described the pattern on Tuesday: new technology layered on top of old systems, and more of these failures to come as the upgrades continue. He also suggested that companies warn customers before scheduled upgrades. We agree with him on both points.
What it means for your business
Your work now depends on systems you do not control. When Microsoft, Apple, or your bank goes down, the fix has to come from them, and your best move is usually to wait. That will keep happening. The bad weeks will not always be this crazy, but they will come.
The businesses that came through this week calmly had three things in common. They knew how to reach each other without email. They did not make changes to their own systems in a panic. And they knew where to find out what was going on.
What we recommend
Agree on a backup communication channel. Decide now how your team reaches each other when email is down: a phone list, a group text, or Slack chat. Keep the phone numbers of your key clients and vendors somewhere other than your email.
Sit tight during someone else's outage. Changing passwords, reinstalling apps, or resetting accounts will not bring a provider back, and it can slow your recovery once it returns.
Be suspicious of outage-themed messages. Attackers follow the headlines. During the Microsoft outage, part of Microsoft's own protection against malicious links was degraded. If a message offers to fix an outage or asks you to sign in, call us first.
Claim what you are owed. Microsoft owes credits for their outage, and they are not automatic. We are filing claims for every Microsoft 365 environment we manage.
Decide what can wait. Most work can pause for a day. Name the few things that cannot, such as a filing deadline or a payroll run, and agree now on how they get done without email.
How we handle change on our side
The lesson of this week applies to us too. Our changes to your systems follow a rule: when a change could disrupt your work, such as a firewall update, we schedule it with you in advance. When an urgent security fix carries almost no risk of disruption, we act first and tell you afterward, because waiting would leave you exposed.
One place to check
When a major provider has an outage, feedwire.com/status will carry our current read and our advice. Our record of the Microsoft outage, including how to think about service credits, is there now.
Questions: +1-310-984-6946 or [email protected]
Further reading
CNN on the NSA network outage: https://www.cnn.com/2026/11/17/politics/highly-classified-national-security-communications-network-goes-down
Reuters on the Homeland Security Secretary's comments: https://lufkindailynews.com/news_reuters/business/us-official-says-recent-tech-woes-not-tied-to-cyberattacks/article_4b1c27f8-c996-53d1-b5e7-831efca00294.html
ABC7 on the bank outages: https://abc7news.com/live-updates/bank-outages-today-live-updates-chase-wells-fargo-america-others-are-reported-down/20045306/
Engadget on the Apple outage: https://www.engadget.com/2319504/icloud-and-other-apple-services-are-down-in-an-apparent-outage/
SecurityWeek on the Microsoft outage: https://www.securityweek.com/microsoft-365-outage-disrupts-outlook-exchange-online-and-teams/