Introduction
Last week within about 36 hours of each other, Microsoft 365 stopped working across more than seventy countries, and the FAA halted departures from every commercial airport in the United States.
When large systems fail close together, the instinct is to start looking for an attacker.
But in both of these cases, there wasn’t one. What the two incidents share is something far more ordinary (and for most businesses, far more useful to understand.)
So, let’s look at what actually caused these significant outages, why that cause matters less than most continuity plans assume, and what kind of disruption your business should be preparing for as a result.
What Actually Happened?
Microsoft’s outage began on November 12 and took roughly two days to fully resolve. Exchange Online, Teams, and Copilot were all affected, with mailboxes hit hardest. Microsoft attributed it to configuration settings applied to servers following an update, which overwhelmed the process responsible for establishing secure connections.
Servers stopped handling requests, and the fix involved rolling the change back. Microsoft stated specifically that the outage was not caused by a cyberattack.
The FAA ground stop happened the following day and lasted a few hours, halting around 3,500 departures and producing close to 8,000 cancellations. According to the FAA’s own account, its legacy equipment became unstable when data began flooding the agency’s network at roughly a thousand connection attempts every four to five seconds. The source was traced to a radio communications site in Arkansas where the FAA’s network provider had been deploying a new router as part of routine infrastructure work. The FAA’s statement was direct: based on all available evidence, no indication of a cyberattack or an outside nefarious actor.
When you read those two descriptions next to each other, the resemblance is hard to miss.
The Pattern Is Change, Not Malice
In both cases, somebody did planned, sensible, well-intentioned work. An update was applied. A router was installed as part of upgrading old copper to fiber. Neither decision was reckless, and neither organization was neglecting its infrastructure. The opposite, really: both incidents happened because someone was maintaining and modernizing complex systems.
What followed in both cases was the same mechanism. A new element interacted with existing infrastructure in a way nobody predicted, something got overwhelmed, and the failure cascaded outward far past the component that started it.
This is how complex systems actually break most of the time. Not through a dramatic breach, but through a change that was correct in isolation and unpredictable in combination. We’ve written before about the gap between what you configured and what you think you configured, and this is that same problem at national scale.
Your Customers Won't Ask Why
Here’s the part that matters for planning: Your customers won’t care what the cause behind the outage was. the result for them is the same either way.
If your business had been unable to send email for two days last week, your customers experienced exactly what they would have experienced during a ransomware incident. Orders didn’t get confirmed. Questions went unanswered. Deadlines slipped. The explanation available to you, that it was your vendor’s configuration change rather than an intruder, would not have improved a single one of those conversations.
Business impact is indifferent to cause. Most continuity planning is not. Plans tend to be organized around attack scenarios because those are what security conversations focus on, which leaves a sizable gap where the far more common failure modes live.
The Dependency You Haven't Mapped
The Microsoft outage exposed something specific that’s worth checking in your own environment.
A great many businesses run email, chat, file storage, and single sign-on through the same provider. When that provider has a bad day, those don’t fail as four separate problems. They fail as one, simultaneously, and companies discovered last week that their plan for communicating during an incident ran entirely on the platform that was down.
That’s a concentration question, not a security question in the usual sense, and it’s one almost nobody has examined. If your team can’t reach each other, can’t access files, and can’t authenticate into anything, the sequence of what you do next needs to have been decided beforehand. It cannot be worked out in the moment, because working it out requires the tools that aren’t available.
When There's Nothing for You to Fix
This category of incident has a feature that makes it genuinely different from a breach: there is no remediation available to you. Nothing in your environment is broken. No patch applies. You are waiting on somebody else, with no visibility into their progress and no influence over their timeline.
Preparing for that looks different from preparing for an attack. It means knowing which of your revenue-critical processes have a manual fallback and which simply stop. It means having a way to reach your own team that doesn’t depend on your primary platform. It means deciding in advance who tells customers what, and how quickly, when the honest answer is that you don’t know when service will return. And it means understanding which single providers, if they failed, would take several of your functions down at once.
None of that is expensive. It mostly requires somebody to sit down and work through it before they need it, which is the same reason it usually doesn’t happen.
Final Perspective
Last week was a useful reminder that the biggest threat to your operations in any given year is probably not an adversary. It’s the accumulated fragility of systems you depend on and don’t control, disturbed by somebody’s reasonable “Tuesday afternoon maintenance”.
The businesses that handled last week well weren’t the ones with better security tools. They were the ones who had already worked out what to do when a vendor goes dark, because they’d thought about it on a quiet day rather than during the outage.
At Secutor, our work on business continuity and incident readiness covers exactly this, including tabletop exercises that walk leadership through the decisions before the pressure is real. If your continuity plan assumes the cause will be an attacker, it’s worth pressure-testing against what actually happened last week.
The cause of your next outage is largely out of your hands. What your business does in response is not.
Connect with an Expert for a Free Consultation
Secutor is your team of world-class problem solvers with vast expertise and experience delivering complete solutions keeping your organization protected, audit-ready, and running smoothly. Use the form below to contact us for a free consultation.

