Skip to content
audiotale.net
Sign in
NewsNews

FAA nationwide ground stop 13 Nov: L3Harris router upgrade blamed, 8,200 flights cancelled

A router installed during a copper-to-fiber upgrade flooded the air traffic control network, officials say; the root cause is still under investigation.

audiotale.net team

· 7 min read

Written by AI from the sources linked below. How we use AI

On this page
  1. Key points
  2. What happened
  3. The explanation, and where reports differ
  4. Background: a network being rebuilt while it runs
  5. What other claims were made
  6. How a router can ground a country
  7. What to do
  8. What we don't know yet
  9. What to watch next
  10. Sources

On Friday 13 November 2026 the US Federal Aviation Administration stopped every departure from every US airport for about five hours, only the third nationwide ground stop it has ever ordered. On Saturday, Transportation Secretary Sean Duffy said the trigger was a router installed during a routine copper-to-fiber upgrade at a remote facility, run by the FAA's telecommunications contractor L3Harris. Whether the router was faulty or was installed or configured wrongly is still open. For anyone who builds or runs networks, it is a textbook case of a small change at one edge taking down a system that thousands of people depend on.

Key points

  • The FAA's nationwide ground stop began at 3:36 pm Eastern on 13 November and was lifted at about 8:35 pm Eastern, according to Ars Technica.
  • Cirium counted more than 8,200 flights cancelled on Friday evening and at least 3,000 more on Saturday, TechCrunch reports.
  • Duffy said a newly installed router flooded other routers with more than 1,000 connection attempts every four to five seconds, disrupting radar feeds and communication with aircraft, according to The Verge.
  • The FAA's acting chief technology officer, Rebecca Guy, said the agency disabled the two sources of the traffic and stabilised the network. A root cause analysis is under way.
  • FAA administrator Bryan Bedford said the agency sees "zero evidence of anything nefarious and zero evidence of anything AI-related."

What happened

At about 3:36 pm Eastern on Friday the FAA issued a nationwide ground stop. Duffy announced it on X, citing a "communication issue impacting all Air Route Traffic Control Centers", the facilities that handle aircraft between airports. Homeland Security Secretary Markwayne Mullin said it was a telecommunications outage inside the FAA and that there was no sign of an attack, according to Ars Technica. The Register reports that the FAA's notice to operators ordered all airports ground stopped, with international inbound flights handled case by case, and that a later update called it "a temporary telecommunications outage". Aircraft already in the air kept flying and landed; the stop applied to departures. San Francisco International Airport told the Register that planes en route were still landing there.

Departures resumed at about 8:35 pm Eastern, Ars Technica reports. By then the disruption had spread beyond the US. Air Canada cancelled all flights to and from the United States for the rest of the evening. Ars Technica says more than a million travellers were affected. Cirium, an aviation data firm, counted more than 8,200 cancellations on Friday evening, and TechCrunch reports at least another 3,000 flights cancelled on Saturday as the disruption carried over.

The explanation, and where reports differ

Friday evening brought a first account. The Register relays a CBS News report that an update to the private network connecting the air traffic control system caused it to reboot, and that when the FAA Telecommunications Infrastructure (FTI) came back up, it crashed. FTI is operated by L3Harris and, according to the Register, connects more than 4,400 facilities.

On Saturday, at a press conference, Duffy gave a more specific account. He said copper wire at a "remote FAA facility" was being replaced with fiber optic cable during "routine maintenance", because the old line was being discontinued. After the work, he said, a new router was flooding other routers in the air traffic control network. The Verge quotes him as saying there were "over 1,000 connection attempts every 4 to 5 seconds". Guy told the same press conference that the FAA found two places pinging its network at that rate, disabled both right away and then stabilised the network, in TechCrunch's report.

These accounts are compatible (a reboot can follow a flood of traffic), but they are not the same, and the FAA has not published a technical report. Treat the router explanation as the government's current account, not a final finding.

Duffy was explicit that cause and blame are unsettled. TechCrunch quotes him as saying the agency is trying to determine "whether we had incompetence", meaning an error by L3Harris, "or did they install a faulty router?" L3Harris has not accepted either description. In a statement given to the Register and The Verge, it said it has completed more than 70 percent of the planned conversion from legacy copper wires to fiber, and that it is supporting the FAA's investigation into the conditions that led to the ground stop.

Background: a network being rebuilt while it runs

FTI is the backbone that links towers, radar sites and control centres. In an 11 May 2026 post, L3Harris said it had passed 50 percent of a nationwide rebuild, was deploying 150 to 200 new network paths a month, and used a "build-then-switch" method: install the new infrastructure while the old system runs, test it in parallel, then cut over in carefully coordinated low-traffic periods with full backup systems in place. Six months later the company says the conversion is above 70 percent.

That pace matters. At 150 to 200 new paths a month, the programme is making changes to the live network all the time, and each one is a chance for a configuration mistake or a hardware fault to reach the live network. Asked why this switchover happened at 3 pm on a Friday and not at midnight, Duffy said switchovers at big airports happen late at night but remote ones happen in the daytime, and that "there will be reevaluation of when we do this work", according to TechCrunch.

Before Friday there had been two nationwide ground stops: the one after the attacks of 11 September 2001, and a short one on 11 January 2023 when the Notice to Air Missions (NOTAM) system failed, Ars Technica notes.

What other claims were made

The outage came a day after the FAA announced a wider rollout of an AI tool for managing air traffic congestion, and the air traffic controllers' union said it had not been consulted on that system, TechCrunch reports. Bedford rejected any link. Air Space Intelligence, which holds the FAA's $875 million, 12-year contract for its SMART airspace system, also said in a statement carried by the Register that its products played no role in the ground stop. We have found no evidence in any source that AI was involved, and the officials say there is none.

Duffy used the episode to press Congress for money, saying lawmakers had provided only a "partial payment" for the equipment work. On X he blamed "DECADES-OLD analog equipment" and said there would be "more of these" without major updates. That is a policy argument. The stated cause on Saturday was a new router on a fibre upgrade, not old equipment failing on its own, and readers should keep those two claims apart.

How a router can ground a country

The details we have point to a familiar failure pattern: a broadcast storm or connection flood, in which one misbehaving device generates traffic that overwhelms others. A rate of more than 1,000 connection attempts every four to five seconds is not large by internet standards. It matters here because these links carry radar and voice data that controllers need in real time, and the FAA chose to halt departures nationwide rather than fly without that data (our reading of the reports; the agency has not explained its decision in detail).

We do not know the router's make, its configuration, what protocol the flood used, or why safeguards did not contain it to one site. Those are the questions an investigation should answer. Without them, the lessons are general ones that operators of any critical network will recognise: limit how much traffic any one new device can send into the core, test cutovers against the real failure modes, keep a fast way to isolate a site, and schedule changes with the cost of a failure in mind, not the convenience of the crew.

What to do

  • Travellers: check your airline's app or site for rebooking before going to the airport; carriers were still cancelling flights on Saturday.
  • Network and platform teams: review how a single new device at an edge site can affect your core. Check that you can disconnect one site quickly, and whether your change windows reflect the damage a failure could cause.
  • Everyone else: nothing needs patching. This was an operations failure, not a vulnerability.

What we don't know yet

  • The router's manufacturer, configuration and the nature of the traffic it generated.
  • Whether the cause was a hardware fault, a configuration error or an installation mistake. Officials have named all three as possibilities.
  • Why the failure at one remote site spread to all the Air Route Traffic Control Centers.
  • How the FAA's Friday-night account (a reboot and crash of FTI) fits with Saturday's account (a flood of connection attempts).
  • Whether the FAA will publish its root cause analysis, and when.

What to watch next

  • The FAA's root cause analysis and any report to Congress, including who bore responsibility under the contract.
  • Whether the FAA changes the timing and approval process for switchovers.
  • Funding for the remaining conversion work, which Duffy has tied to the outage.
  • Whether L3Harris says more about its method after the investigation.

Sources

Written by

audiotale.net team

The official account of the people who run audiotale.net: announcements, conversation starters and the newsroom's drafts. Posts written with AI say so. Every other account here is a real person.