---
title: Operational resilience in banking: Lessons from the FAA ground stop
source_page: https://www.sardine.ai/blog/operational-resilience-in-banking
canonical: https://www.sardine.ai/blog/operational-resilience-in-banking
format: text/markdown
date: 2026-11-18T14:00:00.000Z
author: Hailey Windham
description: Operational resilience in banking is a fraud problem too. The FAA ground stop shows how one problem in a connected network can stop a national system…
---

**Quick links:** [Human page](https://www.sardine.ai/blog/operational-resilience-in-banking) · [Home](https://www.sardine.ai) · [Customers](https://www.sardine.ai/customers) · [Blog](https://www.sardine.ai/blog) · [Demo](https://www.sardine.ai/demo)

---

# Operational resilience in banking: Lessons from the FAA ground stop

**Author:** Hailey Windham · **Published:** 2026-11-18T14:00:00.000Z · **Read time:** 5 · **Categories:** Fraud, Community Banks, Credit Unions

Operational resilience in banking is a fraud problem too. The FAA ground stop shows how one problem in a connected network can stop a national system…

The FAA grounded flights nationwide on November 13 over a telecommunications fault, and banking runs on the same kind of interconnected, aging infrastructure, which makes operational resilience a fraud problem and not only an IT one.
What the ground stop means for banking:
One problem in an interconnected network grounded US flights. Banking runs on the same kind of aging, interconnected infrastructure, which makes operational resilience in banking a fraud problem and not only an IT one.
Shared banking infrastructure has failed before. The Federal Reserve’s February 24, 2021 disruption was operational error not a fraud scheme, and backlogs remained after services resumed.
When visibility changes, a fraud team’s ability to detect suspicious activity can change with it. Teams need to know whether alerts still generate, whether transactions are held or queued, and whether their manual procedures have been practiced.
Multiple vendors do not guarantee independent systems, because vendors can share the same underlying infrastructure and a fraud team has to know where that overlap sits.
Fraud teams are already stretched. In Sardine’s benchmarking survey reporting teams are strained or over capacity, which is why an outage plan needs to settle in advance who can restrict a payment channel and what would justify it.
What happened on November 13
On Friday, November 13, the FAA's telecommunications network became unstable and the agency issued a nationwide ground stop. It lasted more than four hours, with air traffic resuming around 8:15 that evening. In its explanation, the FAA said legacy equipment became unstable when the network started receiving about 1,000 connection attempts every four to five seconds. The problem affected 17 of its 22 Air Route Traffic Control Centers.
The initial investigation pointed to two possible locations flooding the network with excess data. One was in Arkansas, where a contractor was doing routine telecommunications work, including installing a new router, as part of a move away from aging copper-based infrastructure. The other was a radar site in Amarillo, Texas, where the same contractor had completed a similar upgrade about three weeks earlier.
So far, investigators have focused on the upgrade work itself. The FAA has said it has no indication this was a cyberattack, and the investigation is still ongoing.
That's why I built this week's episode of Fraud Forward around it:
"A problem that began within an interconnected technology environment affected an entire national system. One problem. Nationwide consequences. That's the part that should make every financial institution stop and think."
Banking infrastructure has its own single points of failure
We have already seen shared financial infrastructure stumble. On February 24, 2021, the Federal Reserve had a service disruption that affected Fedwire, FedACH, and other payment services. The Fed traced it back to an operational error, not a fraud scheme. While the outage lasted only a few hours, significant backlogs and transaction-processing issues remained after services resumed.
The lesson for fraud leaders is about control:
"You can have the best fraud team in the world. You can have incredible technology. You can have policies, procedures, controls, monitoring, reporting, and all the things we spend our careers building. Yet your institution still relies on infrastructure you don't completely control."
Becky Reed, a financial services strategist who’s spent 30 years in the credit union space, named the same weak point when I asked her where the system is most fragile on our 100th episode: "The legacy antiquated systems we're all forced to use." She described the pain of "fragmented, batch-focused processes that are reactive as opposed to being proactive."
Becky and I have been circling this for years. In episode #6, Fraud Detection Technology and Blockchain in Credit Unions, we talked about how traditional payment systems still rely on delayed reconciliation and fragmented data sharing, and what shared, tamper-resistant records could mean for investigations. The point we kept returning to applies here too:
"Technology alone does not solve fraud. Strong governance, collaboration, and education are still the foundation."
Vendors add another layer. Back on episode #100, Becky argued for orchestration over patchwork, saying the response should happen in "a single platform, not 25 different vendors trying to bolt on to a system that's 50 years old." I come at that patchwork from a different direction.
"Having multiple vendors doesn't automatically mean you have multiple independent systems. They may share the same underlying infrastructure. That's not necessarily a problem. It becomes a problem when we don't know it."
Ironically, we got another example on Monday, November 16, when a software bug slowed transaction processing at JPMorgan Chase. Chase acknowledged the delays may have reached other banks receiving its transactions. By that afternoon, customers at Bank of America, Wells Fargo, U.S. Bank, and Truist were reporting problems too. U.S. Bank warned clients that Zelle payments with other institutions could run late, even though its own systems weren’t directly affected. A Chase spokesperson ruled out a cyberattack.
Why an outage is a fraud problem
Picture a community bank or credit union on a Friday afternoon when one of those dependencies goes down, whether it’s the core, the payment processor, or the integration between the fraud platform and transaction processing. Customers are calling, branches are fielding questions, the contact center is overwhelmed, and IT is racing to get systems back up. Meanwhile the fraud team has to work out what it can still see.
"Can you still see the activity you normally monitor? Are alerts still generating? Are transactions being held? Are they being queued for processing?"
Then comes recovery, which raises its own set of questions: do alerts generate retroactively, and how do you sort out what was authorized, processed, failed, or needs another look? I’m inclined to call these "fraud questions," not IT questions, because "when visibility changes, your ability to detect suspicious activity may change with it."
Manual procedures can fill the gap if they are established, controlled, and practiced. Without that, staff end up making decisions without the information they normally have.
There is also the criminal's view of the same event. I'm not saying every outage leads to fraud, or that the FAA incident was connected to fraud. But I do worry about who’s paying attention while everyone else is scrambling.
"Fraudsters don't necessarily have to cause a disruption to take advantage of the confusion surrounding it."
They could use a disruption as a pretext to contact customers, impersonate the institution and ask members to "verify" accounts, or exploit confusion about delayed payments and account access. We've seen criminals exploit emergencies, natural disasters, and major news events, and now increasingly technology disruptions belong on that list.
Fraud teams are already stretched
Sardine's Financial Institution Fraud Benchmarking Report surveyed 165 fraud leaders at community banks, regional banks, and credit unions. Three findings matter most for this conversation:
71.1% said their teams were strained or operating over capacity.
84% said they were satisfied with their fraud technology.
62.3% still named high false positives as a limitation, and respondents also raised concerns about configuration flexibility, core integrations, and reporting.
Now add an outage to a team in that position. They are answering leadership, briefing branches, calming customers, checking whether controls still work, and reviewing backlogs once systems return.
"That's a LOT to put on a team that's already struggling with capacity."
None of this is an argument against new fraud technology. I believe we need better technology, and I believe AI has real potential to help these teams. But a new tool does not make a program resilient on its own:
"What good is an incredible fraud detection system if the data it relies on stops flowing? What good is real-time decisioning if your integration isn't functioning?"
Becky made a related point on episode #100 from the other direction. In her view, institutions that hold back on newer technology leave themselves exposed: "if we stick our head in the sand, we don't understand what this new technology means ... then we are going to be victims of circumstance by allowing the future to pass us by." Resilience and innovation have to be built together.
Who has the authority to hit the brakes?
The FAA decided the system was not stable enough to keep flying, and it stopped, despite stranded passengers and lost revenue. That raises the question:
"What does a ground stop look like in banking?"
If your institution found a technology issue that could compromise transaction monitoring or payment integrity, could it restrict a payment channel, suspend certain transactions, or require extra verification? Are there thresholds for when to do that, and who makes the call: fraud, IT, risk, or executive leadership? And what happens when the person who normally decides is unavailable?
Banking is not aviation (obviously). There are legal, operational, liquidity, and customer considerations, and shutting down payment systems at every hiccup would create its own problems. The goal is to decide in advance what would justify restrictions, who can approve them, and how they would be communicated.
"Because the middle of a crisis is a terrible time to discover nobody knows who gets to make the call."
That is where tabletop exercises and documented procedures earn their keep. A business continuity plan means little if the fraud team has never tested it against questions like what happens to alerts, delayed transactions, and customers being targeted by scams during the disruption.
Recovery is its own phase of work. The FAA's flights resumed Friday evening, but that didn't mean airlines were instantly back to normal. For financial institutions:
"Restoring a system is one thing. Restoring normal operations, reconciling transactions, validating controls, and rebuilding confidence can be something entirely different."
This is also why fraud leaders need a seat at the table for resilience, technology modernization, vendor relationships, and business continuity. Being valued is different from being involved in the decisions that shape your institution's risk.
Five questions to take back to your organization
Instead of focusing only on the next fraud threat, look at the systems you depend on to fight it. Start here:
Do you know which fraud controls depend on your core or payment processor? Know what happens if the integration stops. Do alerts stop? Do transactions queue? Are decisions delayed?
Do you understand your critical vendor dependencies? Multiple partners don't guarantee independent infrastructure. Find where services overlap and whether your backup plan protects you the way you expect.
Does your business continuity plan account for fraud? Not just cybersecurity and technology recovery. How are suspicious transactions identified during an outage, how are unusual customer requests handled, and what extra verification is needed?
What happens when everything comes back online? Who reviews activity from the disruption, how are missed alerts found, how are backlogs managed, how do you prevent duplicate processing, and how do you know your controls work again?
Does your fraud team have a seat at the operational resilience table? If not, start that conversation. You don't have to own the whole continuity program to contribute to it.
Innovation and operational resilience have to be built together
I don't know whether a nationwide banking ground stop is coming, and I certainly hope not. Our financial system has safeguards, redundancy, and resilience measures. The open question is whether they are keeping pace with how financial services are changing, because "the speed of innovation isn't slowing down. The speed of money isn't slowing down. Fraudsters certainly aren't slowing down."
So here is the challenge I’m giving listeners during International Fraud Awareness Week: ask one question your organization may not have considered. Find out what happens when a critical system fails, whether your fraud controls would still work, and who has the authority to pause activity.
"Innovation without resilience isn't progress. It's risk we may not fully understand yet."
Hear the full conversation in this week's Fraud Forward episode, "One Outage Away: The Fragility of Modern Banking," out Wednesday, November 18.
Keep listening:
Episode #6: Fraud Detection Technology and Blockchain in Credit Unions, with Becky Reed (June 12, 2024)
Episode #100: Payment Fraud Prevention and the Future of Trust, with Becky Reed, Karisse Hendrick, David Maimon, and Jeff Taylor (May 20, 2026)
Answers to FAQs to have before the next outage
### What is operational resilience in banking, and why is it a fraud problem?
Operational resilience in banking is the ability to deliver operations, including critical operations and core business lines, through a disruption from any hazard, according to the interagency sound practices paper. That paper is written for the largest firms, and the Federal Reserve says operational resilience matters to firms of all sizes. It becomes a fraud problem because an outage changes what a fraud team can see. Alerts may stop generating, transactions may be held or queued, and staff may have to decide without the information they normally rely on.
### What caused the FAA ground stop, and what does it teach banks about single points of failure?
The FAA said legacy equipment became unstable when its telecommunications network began receiving about 1,000 connection attempts every four to five seconds. The instability affected 17 of its 22 Air Route Traffic Control Centers and led to a nationwide ground stop on November 13. The FAA has said it has no indication of a cyberattack, and the investigation is ongoing. For banks, the lesson is that one problem inside an interconnected environment can reach a whole system. Fraud leaders should know which of their controls depend on the core, the payment processor, or the integration between them.
### Has shared banking infrastructure failed before?
Yes. On February 24, 2021, the Federal Reserve said an operational error disrupted services including Fedwire Funds and FedACH for several hours. It later traced the error to an automated maintenance process that was inadvertently triggered during business hours. After services came back, the Fed told banks that a backlog could take time to clear and urged them to reconcile messages before sending more, which shows that restoring a system and restoring normal operations are different jobs.
### What should a fraud team’s incident response cover when the core or payment processor goes down?
It should start with visibility. Can the team still see the activity it normally monitors, are alerts still generating, and are transactions being held or queued? Recovery then raises its own questions, including whether alerts generate retroactively and how to sort out what was authorized, processed, failed, or needs another look. Manual procedures can fill the gap only if they are established, controlled, and practiced. Criminals can also use the confusion, so expect impersonation attempts that ask customers to “verify” accounts or exploit delayed payments.
### Does using multiple fraud vendors make a bank more resilient?
Not automatically. Multiple vendors can share the same underlying infrastructure, which makes them less independent than they look, and that becomes a problem when the institution does not know it. Map where services overlap and test whether your backup plan protects you the way you expect. Becky Reed argued on Fraud Forward episode #100 for orchestration in a single platform instead of many vendors bolted onto an old system, and the interagency sound practices also ask institutions to identify replacement third parties for critical services.
### Who should have the authority to pause payments, and why does the fraud team need a seat at the resilience table?
Decide in advance who can restrict a payment channel, suspend certain transactions, or require extra verification, what thresholds trigger that decision, and who acts when the usual decision-maker is unavailable. Tabletop exercises and documented procedures test whether a business continuity plan accounts for alerts, delayed transactions, and customers targeted by scams during a disruption. Fraud leaders belong in that planning because the sound practices tie operational resilience to governance, business continuity management, and third-party risk management, and Sardine’s benchmarking page notes that fewer than half of institutions share fraud data with their AML, cyber, or enterprise risk teams.
