Welcome to Staten News, where the inbox died first, the planes stopped second, and the spy agency's network went down on day six.
Between Nov. 12 and Nov. 17, nine separate organizations had major disruptions: Microsoft 365, Anthropic's Claude, Workday, the FAA, AWS, JPMorgan Chase, United Airlines, the NSA and Apple. TechRepublic put the whole week into one timeline on Wednesday. It reads like a stress test nobody scheduled.
Everybody wanted it to be an attack. So far, nobody has found one.
🧯 The Week, in Order
Microsoft went first. Exchange Online was down from 4:20 p.m. UTC on Nov. 12 to 4:10 p.m. UTC on Nov. 13, just under 24 hours, and Teams, Copilot, SharePoint and OneDrive features that lean on Exchange broke with it. Microsoft's preliminary report blamed a software defect introduced during an earlier network capacity upgrade.
Nov. 13 got crowded. Workday customers lost HR and payroll after a disaster-recovery problem left inconsistent production data. Claude had two separate sign-in and error incidents across Claude.ai, Claude Code and the API, lasting about two hours and two and a half hours. Then the FAA called a nationwide ground stop from about 3:30 p.m. to 8:15 p.m. ET after its legacy network got hit with roughly 1,000 connection attempts every four to five seconds, knocking out communications and radar feeds at 17 of its 22 en route centers.
The weekend didn't help. AWS had authentication and connectivity errors in US-East-1 for about three hours on Nov. 15. Chase delayed payments and some Zelle transfers on Nov. 16 because of a software bug in an on-premises system. United pilots lost access to flight planning and final aircraft weights the same day. On Nov. 17, the NSA reported degraded internal systems during scheduled maintenance, and Apple said an internal update knocked out parts of Apple Music, the App Store and iCloud.
Translation: the most common cause of a catastrophic outage this week was somebody doing routine work.
🔗 The Dependency Problem Nobody Prices In
Microsoft says it found no sign of malicious activity. Chase says the same. The FAA says it has no indication of a cyberattack. AWS, United and Claude haven't published a definitive root cause, and that's exactly the gap the conspiracy crowd has been living in all week.
The more useful story is how the failures traveled. Claude's Microsoft 365 connector and Claude for Outlook threw errors because Exchange was down, which Anthropic confirmed separately from its own incidents. AWS's IAM Identity Center hiccup took services that depend on it, like Amazon Connect and Kiro, along for the ride. Apps were running fine. People just couldn't log in to reach them.
That's not a cybersecurity story. That's a concentration story.
Every enterprise sold "multi-cloud" over the last five years. Plenty of them still run every login through one identity provider and every email through one tenant. Redundancy on the slide deck, single points of failure in production.
🏢 Who Pays for a Week Like This
The incumbents get the bill and keep the customers. Nobody is migrating off Microsoft 365 because Outlook was down for a day, and nobody leaves US-East-1 over three hours. Switching costs are the moat, and outages don't drain it.
The winners are the boring vendors: observability, incident response, backup identity, anything that sells "what happens when your provider goes dark." Those budget conversations get a lot easier after the CFO's payroll run fails.
And the timing for Anthropic is awkward. Claude's two outages landed six days before the company went public on Thursday. A brand-new public company with paying enterprise customers now gets asked about uptime on every earnings call, not just model benchmarks.
📅 What Comes Next
The FAA is still working on a final root cause, and investigators have flagged two possible sources of the traffic flood, including an Arkansas site where contractor L3Harris had been installing a router. AWS, United and Apple still owe the public a real postmortem.
Because when nine systems fall over in six days and the explanation is "maintenance," the next question isn't who attacked them. It's who's checking the maintenance.
This is the week enterprise IT found out how many of its backup plans quietly share the same login screen.
Sources: TechRepublic November 2026 outage timeline, Fox Business on the Anthropic debut
— The Bandicoots 💻🔥

