Field report NE-36 / M5 · sheet /operations

Operations & RestructuringFiled from the territory

Four Unrelated Failures in Five Days Hit Email, Flights and Payments

Microsoft, the FAA, Chase and United each failed separately within five days, with no cyberattack found. One router halted 12,462 flights, and its failure cause remains unknown.

Grid
NE-36 / M5
Surveyed
November 20, 2026
Reading
5 min read
Correspondent
Sophie Lindqvist
Microsoft, the FAA, Chase and United each failed on their own in one week
Fig. NE-36 / M5Microsoft, the FAA, Chase and United each failed on their own in one week — falco / pixabay

Field notes

  1. Microsoft 365 crashed worldwide for about 24 hours on Nov. 12 from a latent defect in a July update to a networking gateway.
  2. A single router in southern Arkansas halted all U.S. departures for under five hours; about 12,462 flights were canceled.
  3. United's Nov. 16 outage blocked pilot pre-flight applications, delaying departures roughly an hour; the fault took 1 1/2 hours to resolve.
  4. A Chase software bug on Nov. 16 slowed transactions at rival banks and on Zelle; no affected-customer count was disclosed.
  5. Duffy asked Congress for $50 billion for air traffic control upgrades; the failed router's cause is still unknown.

Four critical American systems failed within five days in mid-November, and every operator involved says it found no cyberattack. Microsoft's email went down first on Nov. 12, the FAA halted every U.S. departure, Chase payments slowed, and United held planes at the gate on Nov. 16, according to PBS News. The pattern looks coordinated. By every official account, it was not.

Homeland Security Secretary Markwayne Mullin says the failures were unrelated to one another. Microsoft ruled out both a cyberattack and its own AI agents as causes, ABC News reports. The FAA and the Department of Transportation saw no sign of an attack, according to CNBC. The Associated Press reports Chase also ruled one out, and United found no indication its problem involved cybersecurity.

That absence should worry executives more than a hacker would. Each breakdown appears to have come from inside systems that institutions trust to keep running. "We need to know the systems that keep this country running won't simply fall over without notice," Prins Ralston, CEO of the Australian Computer Society, told ABC News.

What did the Microsoft outage actually break?

Microsoft 365 failed on Nov. 12, and Audacy reports the fault traced back to an update made in July. That update left a latent defect in a networking gateway, a device that routes traffic between networks. Nobody caught it until services crashed worldwide for about 24 hours. Users in any region lost access, ABC News reports, so one internal fault cut off customers everywhere at once.

Recovery outlasted the crash. Services went down on a Thursday and were not fully restored until Saturday morning, according to ABC News.

The blast radius reached far beyond offices. Immigration New Zealand halted all visa processing, ABC News reports — a national government stopping its own service over a foreign company's network fault. Health staff were "unable to refer patient details in the normal way," Rachel Harrison, public services head at the GMB union, told ABC News. Small businesses lost up to $30,000 in one day, ABC News reports, though they had no part in the fault and no power to fix it.

What stopped every U.S. departure?

For part of one day, no plane in the country could take off. The FAA stopped every nationwide departure for under five hours, and about 12,462 flights were canceled that day, Audacy reports. The halt reached tens of thousands of flights in total, CNBC reports. Every aircraft already in the air landed safely.

Transportation Secretary Sean Duffy linked the stoppage to a single router in southern Arkansas, according to CNBC. The router belonged to an air traffic control upgrade, and L3Harris Technologies is the contractor in charge of that system. One device from the program meant to improve the network stopped flights everywhere.

The cause remains unknown. "Identifying the device that failed does not establish why it failed," Democratic members of the Senate Commerce, Science and Transportation Committee said in a statement, as reported by CNBC. Until someone finds the cause, nobody can promise the same part, installed the same way, will not fail again.

How did a pilot-app outage ground United?

United held departing planes at the gate on Nov. 16 because pilots could not reach the applications they use to prepare for takeoff, Audacy reports. Departures were delayed roughly an hour. The underlying problem lasted longer: PBS News reports 1 1/2 hours passed between the first pilot reports and resolution. A one-hour ground delay did not mean the tools worked again within an hour.

The damage stayed narrow. Flights already in the air continued unaffected, PBS News reports, and United Express regional flights kept operating. The exposure for travelers sits on the ground before the doors close, not after.

Was the Chase slowdown as small as Chase says?

On Nov. 16, a software bug at JPMorgan Chase slowed some transactions. Audacy reports the problem was fixed the same day and was not systemic. Customers kept access to their accounts and money, the Associated Press reports.

The effects crossed bank lines anyway. Bank of America and Citigroup had no trouble with their own systems, the AP reports, yet their customers saw delays whenever a payment passed through Chase. Zelle users were caught too, though a spokesperson for network owner Early Warning Services says Zelle was fully operational. "Other banks may have been impacted by slowness in transactions only to the extent that they were a party to a transaction with us," Chase spokesperson Joe Evangelisti said.

Note what is missing here: no count of affected transactions or customers has been reported. The reassuring account comes from the bank that failed. Chase presented a same-day fix as an expectation rather than a confirmed result, the AP reports. The week's mildest failure is mild by the measure of the one company with the most reason to call it that.

What does the week change for buyers of critical systems?

The common thread is not a shared attacker but shared fragility. Each failure began with a different, ordinary fault, and each spread in its own way — across borders, through an industry, or to customers of rival banks. For some failures, the all-clear rests entirely on the company's own word, and for at least one, the router, nobody yet knows why it happened.

Washington now faces the same bet at scale. Duffy asked Congress for $50 billion for air traffic control upgrades, CNBC reports, and Senate Commerce Chair Ted Cruz doubts lawmakers will approve the full sum. Committee Democrats want new equipment to keep flights as safe as the old system did.

The risk is not only that the old system breaks. It is that the new one goes in faster than anyone understands how it fails — as the Arkansas router already demonstrated.

via qz.com (Source)

Share this article:

More from Sophie Lindqvist

Sophie Lindqvist

Show full bio

Correspondent covering money, funding and market shifts around corporate strategy and chief-executive leadership at CEO Frontier.

495 articles

Filed under

Adjacent sheets

« Previous article

End of sheetNE-36 / M5 · November 20, 2026