A networking outage that saw the Federal Aviation Administration (FAA) order a nationwide ground stop for all U.S. airports cast a spotlight on an all-star cast of network giants including Hewlett Packard Enterprise (HPE), several U.S. telecom giants, and Lumen Technologies, which told this title that aging copper infrastructure is no longer fit for purpose.
The incident appears to have been triggered by a network disturbance during a routine telecommunications upgrade by defense contractor L3Harris Technologies. The investigation has since drawn in HPE and its Juniper Networks networking business, while raising timeline questions about the FAA’s separate replacement-network program involving AT&T, Verizon, and Lumen.
How the FAA ground stop unfolded, and the subsequent blame game
On the afternoon of Friday, November 13, the FAA's telecommunications network became unstable, with router-related traffic reportedly reaching more than 1,000 connection attempts every few seconds, disrupting some aircraft radio communications and radar data feeds. Subsequently, the FAA ordered a nationwide ground stop, which lasted until 8:15 p.m. after stabilizing the system. Tens-of-thousands of flights were affected in the longest grounding of flights since the 9/11 terrorist attacks in 2001.
Rumors circulated that the outage was connected to Microsoft’s global cloud failover the same day, as well as an AI-powered system deployed days earlier by the FAA, known as SMART (Strategic Management of Airspace, Routes, and Trajectories). A possible cyberattack was also mooted.
In a subsequent press conference, Transportation Secretary Sean Duffy identified L3Harris's copper-to-fiber migration work as a potential trigger. L3Harris confirmed it was “modernizing the communications infrastructure supporting the National Airspace System in close coordination with the FAA” and cooperating with its investigation, without accepting responsibility for the outage.
The contractor’s statement claimed that together with its telecom partners and the FAA, it completed more than 70% of the planned airspace communications conversion from legacy copper wires to a fiber optic system.
In the same press briefing with Duffy, FAA Administrator Bryan Bedford revealed that the agency was replacing around 2,000 Cisco 6000-Series routers with devices from Juniper Networks. Reporting from The Wall Street Journal suggested a Juniper Networks device in southern Arkansas was to blame for the entire outage, something Juniper Networks parent HPE strongly refuted.
"Based on what we know today, the device functioned as it was configured to by the installer. It did not fail or act improperly," an HPE spokesperson told ABC News.
Duffy would go on to reveal the suspect router was being examined at an L3Harris lab in Florida, without confirming whether it was a Cisco or Juniper Networks device. A second rogue device has also been identified in Amarillo, Texas.
Roy Chua, founder and principal at research firm AvidThink, advised against pinning the outage on any one vendor at this stage, while drawing parallels with the November 2009 FAA Telecommunications Infrastructure (FTI) outage, for which Harris Corporation was blamed for after the misconfiguration of a replacement router during planned maintenance.
“To me, both events point at change management as much as equipment age,” Chua said.
Referring to the suspected Arkansas and Texas sites, the analyst was strong in his opinion that two edge sites “shouldn't be able to destabilize 17 centers.”
“That's a blast radius problem, and we have best practices on this: segmentation, control-plane policing and rate limits, storm control, turning off features that aren't in use, and out-of-band management. The FCC (Federal Communications Commission) drew similar lessons from CenturyLink's 37-hour outage in 2018. And that's just one example,” Chua explained.
The past and future of FAA networking
L3Harris inherited the FTI when Harris Corporation merged with L3 Technologies in June 2019; the Harris contract dates back to 2022.
With L3Harris maintaining the legacy infrastructure, the FAA has looked to AT&T and Verizon for building its successor network. Verizon won the $2.4 billion FAA Enterprise Network Services (FENS) contract in 2023, only for things to change when AT&T was appointed as the project's new lead contractor. As announced in August 2026, AT&T received an initial $74.3 million planning and development award. AT&T’s appointment came out of urgency with the FAA wanting the new network ready by 2028.
Interestingly, Lumen’s role in the upgrade was not public knowledge until the weekend, when the vendor was named by Duffy as one of the three companies working on FAA’s new network.
“I have AT&T, Verizon, and Lumen all ready to go,” Duffy declared, before bemoaning a lack of funding to help the body overcome legacy architecture, repeating similar remarks from September when major flight delays hit the East Coast after a backup fiber optic cable was severed during construction work.
“One of the biggest problems we continue to see is our network, and the architecture of our network is from 1980. We need modern technology and architecture,” Duffy said. “We have the architecture ready to build a state-of-the-art telecom system that will give us resiliency. All I need is the money to finish it. My hope is if they give me money today, I’ll be going tomorrow to build it and this telecom system would be ready by the end of 2028, the first part of 2029.”
Lumen response calls out copper
No further details were given on Lumen’s role in the upgrade. In response to clarification from this title, a Lumen spokesperson said that Lumen’s network was not involved in the FAA disruption and is operating normally. They also referred to the firm's recent acceleration of its copper decommissioning plans, as revealed in its most recent financial earnings.
"These events made it clear we urgently need to modernize the U.S.’ air traffic control system," Lumen's spokesperson noted. "Decades-old copper networks were never built for today’s demands. That’s why we strongly support the efforts of the federal government to advance the modernization of our networks. Lumen is working to retire its legacy copper and invest in modern network capabilities that mission-critical systems need. We’re working with the Department of Transportation, the Federal Aviation Administration, and our federal partners to advance this work, and we stand ready to help expedite delivery."
Are programmable networks the answer?
Whether a mutivendor strategy can remove network risk, Chua said that while carrier diversity removes single-provider risk, it can’t entirely remove common-mode risk.
“The risk moves upwards to the shared overlay, and to the integrator (or FAA) that's maintaining the layer above," Chua explained. "There has to be a network management system that sits above the physical layers, and that will need to be coordinated by the integrator.”
Chua also noted past efforts to standardize fault reporting via APIs by Mplify (formerly MEF), with the caveat that not all carriers support it.
Such APIs define the network-as-a-service (NaaS) model, with the likes of Lumen and consulting giant EY pushing the idea that programmable networks can mitigate against the risk of outages.
For Chua, such next-generation networks can only partly help against such risks.
“Pre-change validation, staged activation, streaming telemetry, and automated isolation would have helped with detection and containment,” Chua said. “It took about 45 minutes to trace the problem, which the FAA partly attributed to paper maintenance logs. While programmability can speed up containment, it doesn't automatically support it. And as the Rogers (Communication outage) in 2022 and Optus in 2023 showed, change windows and automation are often where these failures start.”
Comments