One bad week isn't a conspiracy. It's a concentration problem
Apple's outage joined airlines, banks, email and the NSA in a rough week, and the pattern points less to hackers than to fragile, concentrated systems.
AI-generated image
If this is all a coincidence, is it a comforting one? Lawmakers from both parties have asked versions of that question this week. Our answer is that the coincidence is the least interesting part. The more worrying part is how ordinary the causes look.
A week of outages, and no sign of an attack
The run started on November 12 with a Microsoft 365 outage that hit Outlook and Teams. It lasted nearly 24 hours. Microsoft executive Charles Lamanna blamed a mix of an undetected software defect, higher traffic and reduced server capacity, according to AOL's timeline. On Friday the FAA halted departures nationwide for several hours. Workday went down the same day. On Monday JPMorgan Chase had transaction problems and United had a systems outage. On Tuesday CNN reported that the NSA's internal network was down, and Apple's iCloud, App Store and Apple Music followed.
Apple said an internal update caused its outage, as Reuters reports, and other coverage says the company ruled out a cyberattack. The NSA called its problem "scheduled routine maintenance". Officials say they see no evidence of an attack, and the reports we reviewed describe no established link between the incidents. We'll take that at face value, with one caveat. A confirmed attack did hit Optimizely, a third-party platform, through malware planted in some clients' code. Nothing we have seen ties it to the other incidents, but it shows the threat is real.
The boring explanations are the scary ones
Look at the stated causes. An internal update at Apple. A software release at Chase. A software defect at Microsoft. Routine maintenance at the NSA. At the FAA, Transportation Secretary Sean Duffy said a router "started to broadcast into our network and started to overload our network", and he also pointed to contractor work replacing copper with fiber. Almost none of this was an exotic attack. These were changes that went wrong.
In our view, that is the lesson. Modern systems change constantly, and every change is a chance to fail. When the system is also a single point of dependence for millions of people, a routine update can disrupt a whole country. Zelle shows the effect. It said its own network was operating normally, yet an issue at the financial institutions behind it may have delayed payments for some customers. That is concentration. Your payment, your flight or your email rests on a stack you can't see, and a failure in one layer can reach everyone.
Cybersecurity expert Alex Hamerstone of TrustedSec warned that ageing infrastructure and increasingly interconnected systems could make technical disruptions harder to prevent. We agree, and we'd go a step further. The NSA said its infrastructure has "multiple layers of redundancy". Redundancy that exists on a slide is not the same as redundancy that works under a bad update. The real test is how fast you roll a change back and how far it can spread before you do.
Hold providers to account, and keep your own plan B
Some of the response so far is unconvincing. President Trump told reporters "Things like that happen" and "It's fine". Neither is a resilience plan. Asking Congress for $50 billion in emergency funding to modernize air traffic control, as Duffy has, may be warranted. But money alone doesn't fix the habit of rolling out changes without a safe way back. We also don't want every outage to become a cyberattack story. That makes it too easy to avoid asking about testing, staging and rollback.
What should readers and companies do? A few practical things:
- Ask for post-mortems. Apple, Microsoft and Chase should explain what failed and why it spread, not just say a bug was fixed. Microsoft's preliminary incident report is a start.
- Keep fallbacks. Know a second way to pay, keep a second email route, and don't tie your only authentication to one account. The National Weather Service advised having multiple ways of receiving critical information, and that applies well beyond weather.
- Watch for scams. Cybersecurity analysts noted that outages are prime moments for phishing texts and fake bank alerts.
- If you run a business, map which single vendors can stop you. Payroll, email and sign-in are the usual suspects.
Verdict
Apple restored its main services the same day, and its impact was small compared with grounded planes. But it fits the pattern. The week doesn't show that someone is attacking US infrastructure. As far as public reporting shows, it shows that our infrastructure can hurt itself with a bad update. That is arguably more fixable, but only if customers, regulators and shareholders demand it. Until then, treat every major service as something that will fail eventually, and plan for the day it does.
No topics yet: start the first one.
More stories
Washington moves to regulate rented compute, and Alibaba feels it first
The new US rules go after cloud access to AI chips, not just shipments. We think that leaves Singapore chasing predictable supply and makes Alibaba's AI cloud bet riskier.
The DRAM collusion suit survived, and AI can't be the alibi
A judge refused to dismiss the case against Samsung, SK Hynix and Micron. The AI boom explains some of what happened to memory, but it shouldn't shield the oligopoly from scrutiny.
Microsoft's outage shows resilience is a business decision
A global Outlook, Teams and Copilot failure is a reminder that bundling mail, meetings and AI under one vendor is a bet, and the board should own it.
United's AI agent ban is about the customer, not safety
United says it's protecting travelers, but its CEO's "another middleman" line gives away the real fight: who owns the relationship. Expect licensing, not a permanent wall.



