The absence of a definitive root cause for the FAA outage left HPE temporarily holding the bag, but experts said that's a hasty reaction and hinted at other possible causes, including rapid modernization, configuration error and AI-spurred patching efforts.
The Trump administration blamed last week's FAA ground stop on old technology. The outage occurred while an FAA contractor was switching telecommunications lines from copper to fiber, according to Transportation Secretary Sean Duffy. "Our routers actually got overloaded with data," Duffy said in a press conference.
The agency, which has been trying to modernize for decades, said Tuesday that it had temporarily halted installation of HPE's Juniper devices. The devices that the FAA quarantined to stabilize the network were circuit-to-packet (CTP) devices, according to a source familiar with the issue. CTP devices operate differently from standard IP routers and are primarily used to bridge legacy analog connections to modern IP connections.
Duffy urged Congress to invest $50 billion in the nation's air transportation infrastructure, even as the FAA took steps to prevent a repeat of last week's ground stop.
HPE, which acquired the network device maker Juniper in a $14 billion deal in 2025, saw its stock dip 3% on Tuesday. "Based on what we know and have been told by others, our equipment was working as configured. The FAA is assessing the configuration and performance of all infrastructure involved to determine the cause," an HPE spokesperson told TechTarget.
It's too soon to say who or what's to blame, said Shamus McGillicuddy, vice president of research at Enterprise Management Associates.
"I doubt this is HPE or Juniper's fault. There's very little info about what the true root cause was, and this administration isn't known for openness or honesty. The market is overreacting," he said.
The FAA ground stop comes amid a slew of other prominent outages at organizations including Microsoft, JPMorgan Chase, the NSA, United Airlines and Apple.
"They're pushing new technology on top of really old computer systems, and it just seems to overwhelm the system," Homeland Security Secretary Markwayne Mullin told CNBC on Tuesday. "We're going to see more of this as new technology rolls out."
Mullin sought to dispel rumors about a cyberattack, citing the age of the FAA's equipment.
"We used to have computers that were more powerful than the technology," he said. "Today the technology is more powerful than the computers."
Microsoft's outage similarly involved network infrastructure struggling with increased traffic. According to the company's preliminary findings, a transition to new Teams networking gateways exposed a latent software defect and compounded existing server capacity problems.
Possible causes beyond a single router
The FAA outage occurred during network maintenance by longtime contractor L3Harris, Duffy said at a Saturday press conference. The FAA has been replacing older Cisco routers with newer HPE Juniper equipment, said Bryan Bedford, federal aviation administrator, at the same press conference.
"We are constantly replacing [Cisco 600] routers with Juniper routers," Bedford said. "This is ordinary course-of-business stuff."
Investigators are examining whether faulty equipment or an installation error contributed to the outage, Duffy said. In a Nov. 14 statement, L3Harris said it was cooperating with the FAA's investigation.
It's unlikely that a single router or network device would cause such a sweeping outage, said Ram Chandra Sachan, network architect and distinguished member of technical staff at Wipro, who spoke with TechTarget in a personal capacity and not on behalf of Wipro.
"Yes, a router malfunction or misconfiguration on it can bring down the entire network behind that router [or isolate it] from the remaining network," Sachan said. "However, there is typically always redundancy in place at important exit and entry points in the network. The chances of both routers and respective uplinks failing are very unlikely."
The legacy equipment became unstable when data flooded the FAA FTI network at a rate of 1,000 connection attempts every four to five seconds, which disrupted aircraft communications, according to the FAA.
L3Harris blocked an Arkansas site where the new device installation was taking place, and the network stabilized. A broadcasting function on the new equipment was also disabled as a precaution. This broadcast (or multicast) feature, which sends disparate inputs to multiple destinations, is not enabled by default. Turning it on requires coding knowledge, not just toggling a switch, according to a source familiar with the issue. Several other installations elsewhere have been configured the same way without issue, said the source.
"In my view, this issue has nothing to do with any specific OEM like Cisco, HPE or anyone else. It's [about] misconfiguration or poorly designed network architecture," Sachan said. "I have worked with major OEMs like Cisco, Juniper before it was acquired by HPE, Aruba, which is part of HPE, Ericsson and Palo Alto. All of them have excellent hardware, software and backend teams to support. The only difference is commercial terms."
In my view, this issue has nothing to do with any specific OEM like Cisco, HPE or anyone else. It's [about] misconfiguration or poorly designed network architecture.

Ram Chandra Sachan
Network architect, Wipro
User error isn't out of the question, said Michele Chubirka, principal security engineer and architect at Red Hat. "I'm a former network engineer. It was probably a CK error (a problem between the chair and the keyboard). Lack of documentation, lack of knowledge regarding a legacy infrastructure, probably a misconfiguration," said Chubirka, making it clear that, without a postmortem, she was speculating. "Routers are fairly simple devices, and it's rare to have a 'faulty' Juniper router. It is common to have bad configs or to create a loop because of plugging hardware in wrong," she said.
A 'perfect storm' of outages
It's also worth considering the broader effects of AI-fueled patching and hasty, top-down modernization on network infrastructure, said Patrick Garrity, a security researcher at vulnerability prioritization tool maker Vulncheck.
"One of my biggest concerns about AI is that resources are quickly shifting toward adopting newer technology while the basics get overlooked. Even though we still have limited details about these outages, I'm worried that organizations might be over-rotating to the latest tech at the expense of availability and reliability. That said, a lot of legacy tech genuinely needs modernization, so it really is a double-edged sword," Garrity said. "Kind of a perfect storm happening right now. We are likely to see more of this."
I'm worried that organizations might be over-rotating to the latest tech at the expense of availability and reliability.

Patrick Garrity
Speaker title
AI has industrialized vulnerability discovery, and the volume of findings is surging. Organizations are scrambling to keep up, which pushes them to roll out patches faster and with less testing than they normally would, said Chris Hughes, vice president of security strategy at Noma.
"Fear of disrupting availability has always been a big reason patch timelines run long. Compress those timelines under pressure, and that disruption risk starts hitting availability -- the 'A' in the CIA triad," he said. "The bank CEOs and NSA telling Congress that AI-driven patches are behind recent outages is an early signal of that [disruption]."
Several of the organizations that endured outages in recent days also underwent significant workforce reductions in the past two years, including Microsoft, JP Morgan Chase, the FAA and United Airlines.
"I'm super curious how much of this might tie back to the shifting of resources to deploying AI, resulting in cascading failures in now under-resourced areas," Garrity said.
Ben Lutkevich is an award-winning writer for TechTarget.