A cloud-connected data center monitors flagged risks across email, air travel, banking, and mobile payments.
Flights grounded, email dead, banking sluggish, then Apple wobbling. Last week's run of outages felt like a bad season of a disaster show. The honest answer to "why is everything breaking?" is less cinematic and more useful. These were separate failures, and nobody has shown they share a cause. They do share a structure: tightly coupled systems with very large blast radii.

A cloud-connected data center monitors flagged risks across email, air travel, banking, and mobile payments. The incidents, kept separate​

New Jersey experts weighed in through the Bergen Record. Cristian Borcea, a computer science professor at NJIT, and Mary Chayko, a Rutgers communication and information professor, were both quoted. The incidents they were asked about:

  • Microsoft 365, Nov. 12–13: A global disruption hit Outlook and other Microsoft 365 services.
  • FAA, Nov. 13: A telecommunications problem triggered a nationwide ground stop.
  • JPMorgan Chase, Nov. 16: A software problem delayed transactions.
  • United Airlines, Nov. 16: An outage delayed departures by more than 90 minutes. The airline said it did not appear to be a cybersecurity problem.
  • Apple, Nov. 17: An internal update disrupted services such as iCloud and Apple Pay. Apple said it was not a cyberattack.
  • NSA, Nov. 17: An internal classified communications network had problems during scheduled maintenance. Mission-critical systems stayed up, the agency said.
  • Marriott, Nov. 17: A technical issue hit its website and Bonvoy app. The company said it was resolved and showed no sign of being cyber-related.

Officials have said they see no indication the incidents are related, and none has been attributed to hostile activity. "No evidence of an attack" is not the same as proof an attack was impossible. Still, the stated causes so far are mundane: updates, maintenance and defects.

Microsoft 365: what actually broke​

Microsoft has since explained its outage in detail. Charles Lamanna, Microsoft's executive vice president of Copilot, Agents and Platform, wrote that a latent defect caused servers to become overwhelmed and unresponsive under certain load conditions. Reporting on his post describes three factors combining:

  • The defect came from a fix prepared specifically for Microsoft 365 networking gateway servers.
  • Teams shifted onto the same network that powers Exchange Online, while about 14 percent of server capacity was offline for unrelated maintenance.
  • Together, the extra load and reduced capacity made the gateway hit the latent bug from July, so requests built up until servers stopped responding.

The failure then fed on itself. Traffic management shifted load from failed servers to those still running, which failed too. At the lowest point, more than 80% of servers were unresponsive. Recovery was slow because Microsoft said its normal rollback, flighting and feature flags couldn't quickly revert a change that had been in place for months. One report says the defect was worse on newer machines with more processor cores, while load testing ran mostly on less powerful hardware.

Some outlets describe this explanation as preliminary. One account says the incident record still listed the root cause as pending as of Nov. 14, so details may be refined. The Microsoft ticket for the event was MO1550821.

Microsoft also told ABC News the outage was unrelated to AI. That matters for Borcea's broader argument, covered below.

The FAA ground stop​

The FAA said the problem began around 3:30 p.m. ET on Nov. 13 and air traffic resumed at 8:15 p.m. Some radio frequencies and radar feeds stayed available throughout. The agency's own Q&A, as relayed in our research notes, adds detail:

  • Legacy equipment became unstable when data flooded the FAA Telecommunications Infrastructure network.
  • Investigators pointed to two possible locations, a radar site in Amarillo, Texas, and a radio site in El Dorado, Arkansas. Network provider L3Harris had been working at El Dorado, including installing a new router.
  • The FAA said it saw no indication of a cyberattack.
  • Instability began at 17 of 22 Air Route Traffic Control Centers.
  • The agency said about 12,462 flights were canceled that Friday, though not all were directly caused by the ground stop.

I could not retrieve the FAA page directly, so treat these figures as the agency's account as relayed to us, and note the investigation was still open as of its Nov. 15 statement. The key point is partial degradation. This was not a total loss of radar or radio, but a conservative, safety-first stop after instability spread.

What the experts say about the pattern​

Borcea's point is structural. Cloud-scale systems are complex and interlocking, and they lean on a small number of providers and equipment vendors. When many organizations use the same product, one defect or change can reach all of them. "The blast radius of this failure was very large," he said.

The Microsoft case fits. A fix aimed at one gateway layer, exposed months later by a traffic shift, ended up touching email, calendars and Copilot across industries.

Why backups don't prevent outages​

A copy of your data doesn't keep services running. Restoring can take hours. A standby system on an older software version can allow a switch back, but dependencies between systems can make recovery messy. This is a general resilience lesson, not a claim that any of these companies lacked backups.

Is AI involved?​

Borcea expects more of this ahead and ties it to AI in general. AI lets people produce code faster, he said, creating risk when testing and review don't keep pace, and automated agents can spread mistakes quickly. He also noted AI can help find vulnerabilities faster, provided people evaluate the fixes. This is his analysis, not an incident finding. Microsoft says its outage was not caused by AI or agents, and none of the other incidents has been tied to AI.

Why it feels so unsettling​

Chayko says people rely on devices emotionally as well as practically. Outages can bring "a profound and often frightening feeling of disconnection." She also argues that clear explanations from companies and government matter, because social media rumor and panic fill any silence. Microsoft's detailed post-incident write-up is a decent example of filling that gap with facts.

What admins should take from this​

The Microsoft Q&A guidance on the Nov. 12 event is practical.

  1. Check Service health. In the Microsoft 365 admin center, go to Health, then Service health. Open the active incident and review all updates for mitigation progress.
  2. Know who can see it. Service Support and Helpdesk administrators can view the page too.
  3. Have a fallback. If the admin center is unreachable, use Microsoft's public service status page.
  4. Turn on notifications. Service health lets you enable incident email alerts under Customize, then Email.
  5. Don't fiddle with clients. Reinstalling Outlook or rebuilding profiles won't speed recovery from an acknowledged service incident.
  6. Report after restoration. If your tenant is still affected after Microsoft marks the incident restored, use Report an issue in Service health.

Beyond that, a few steps follow from the reporting:

  • Map which business functions depend on one provider, such as mail, calendar and meetings.
  • Rehearse communication plans that don't rely on the affected system. Email is a poor channel for announcing an email outage.
  • Test rollback and failover, not just backup restoration.
  • Keep patching. The lesson is to stage and test changes, not to avoid updates.

Preparing as an individual​

Borcea keeps local backups of important files on external hard drives instead of relying only on cloud storage, and he recommends installing security updates when available. Chayko suggests paper copies of key documents and a backup plan for reaching loved ones. A local backup keeps your files reachable during a cloud outage, but it won't bring back the service or your access to online accounts.

Bottom line​

Seven unrelated failures in six days do not make a conspiracy. They do show how much everyday life rides on shared infrastructure, where a latent bug, a maintenance window or a router deployment can cascade. The fix for users and IT teams is the same: understand your dependencies and plan for the day they fail.

 

References

  1. NJ experts explain what recent outages reveal about modern tech - Bergen Record Bergen Record 2026-11-19T10:56:16+00:00
  2. Microsoft identifies issues leading to global Outlook outage - ABC News abcnews.com
  3. Microsoft Finally Explains What Caused the Massive Outlook Outage That Disrupted Email for Nearly 24 Hours mensjournal.com