Nationwide FAA ground stop traced to router flooding network during copper-to-fiber upgrade

The US Department of Transportation has confirmed the cause of Friday's nationwide flight ground stop: routine maintenance gone wrong. Work to replace aging copper cabling with fiber optic line at a remote FAA facility left a router hammering the air traffic control network with connection attempts, disrupting radar feeds and communications with aircraft until engineers could isolate the problem.

What actually broke

Speaking at a press conference on Saturday, Transportation Secretary Sean Duffy said contractor L3Harris, a long-time Federal Aviation Administration partner, had been upgrading copper wire to fiber at a remote site. According to The Verge, Duffy explained that the copper line was so old it was being discontinued and would have been shut off, making the replacement necessary rather than part of a broader modernization push.

The trouble began immediately after the switchover. The Verge reports that the newly installed router flooded other routers in the air traffic control system with more than 1,000 connection attempts every four to five seconds, degrading radar feeds and communication with planes.

FAA acting CTO Rebecca Guy, quoted by TechCrunch, said that when engineers noticed the network becoming unstable they identified two endpoints hitting it at that extraordinary rate, disabled both, and stabilized the system. A root cause analysis is now underway.

Error, hardware, or something else

Duffy was blunt about what remains unknown. According to TechCrunch, officials are still determining whether the failure was user or installer error on the part of L3Harris — he used the word "incompetence" — or simply a faulty router. He characterized the underlying job as straightforward work at a very remote location.

The timing has already drawn questions. Asked why the switchover happened at 3pm on a Friday rather than overnight, Duffy said switchovers at larger airports take place late at night while remote ones happen during the day, and acknowledged that the scheduling of such work will be reevaluated.

The scale of the disruption

TechCrunch notes this was only the third nationwide ground stop of the 21st century. Aviation data firm Cirium counted more than 8,200 canceled flights on Friday evening, and at least another 3,000 cancellations followed on Saturday as the backlog persisted even after the ground stop was lifted.

Not the new AI tool

The outage landed one day after the FAA announced the rollout of an AI tool for air traffic controllers — a rollout the controllers' union says it was not consulted on. FAA administrator Bryan Bedford moved to cut off speculation, saying the agency sees "zero evidence" of anything nefarious or anything AI-related, and that the FAA is dealing with a defective piece of equipment.

A funding fight in the background

Duffy has used multiple press conferences to press Congress for more money to upgrade aviation equipment, and did so again. According to The Verge, he argued the FAA is building quickly but has received only a "partial payment" toward the work needed to modernize the air traffic control system.

Why it matters

A single router at a remote facility — misconfigured, faulty, or badly installed — was enough to ground the entire US airspace system. That is a case study in how brittle critical infrastructure becomes when decades-old telecom dependencies are still in service, and how much rides on ordinary change management: maintenance windows, installation verification, and rollback plans.

It also shows how fast operational failures turn into political leverage. Within a day, a maintenance error was recast as evidence that Congress must fund a full air traffic control overhaul. Whether more money would have prevented this specific incident is a separate question, but the episode is a reminder that infrastructure upgrades are themselves a source of risk, and that the safe path — carefully staged, well-timed switchovers — costs planning discipline as much as budget.