The FAA ground stop on Friday, 13 November 2026 followed a flood of data into the agency’s telecommunications network at a rate of 1,000 connection attempts every 4 to 5 seconds, according to a question-and-answer page the FAA published on 15 November. The FAA’s initial investigation traces that flood to two remote sites, one of them a radio facility in El Dorado, Arkansas, where contractor L3Harris had been deploying a new router.
The FAA Ground Stop Followed a Data Flood That Destabilized Legacy Equipment
The FAA says its legacy equipment became unstable when the excess data hit the FAA Telecommunications Infrastructure network, which disrupted aircraft communications and radar feeds. The trouble began at about 3:30 p.m. ET. Instability appeared at 17 of the FAA’s 22 Air Route Traffic Control Centers and spread into several Terminal Radar Approach Control facilities. The agency issued a nationwide ground stop, and about 3,500 aircraft already in the air landed. Some radio frequencies and radar feeds stayed up throughout, the FAA says, so controllers could keep talking to and tracking aircraft. Air traffic resumed at 8:15 p.m. that evening. TechCrunch weekend editor Anthony Ha reported that this was only the third nationwide ground stop of the 21st century.
Two Remote Sites in Texas and Arkansas Sent the Excess Traffic
The FAA names the two sources as a radar site in Amarillo, Texas, and a radio communications site in El Dorado, Arkansas. L3Harris, the FAA’s network service provider, had been working at El Dorado, including installing a new router. Transportation Secretary Sean Duffy said at a Saturday press conference that the job was a routine move from copper wire to fiber optic cable, and that after the upgrade a router overloaded a critical communication system, TechCrunch reported. Duffy said investigators were working out whether the cause was installer error at L3Harris or a faulty router. FAA acting chief technology officer Rebecca Guy said the agency saw two places hitting the network at a very high rate. L3Harris blocked both locations and the network stabilized. The FAA and L3Harris are still investigating the root cause.
The FAA Disabled the Broadcast Function on About 60 Similar Routers
After isolating the two sites, the FAA disabled the broadcast function on the new routers that originated the traffic. As a precaution it then turned off the same function on about 60 similar routers elsewhere in its system. The FAA says it waited for the network to hold steady before lifting the ground stop. It attributes the instability to its legacy time-division multiplexing system, and it says none of the towers upgraded under its separate modernization funding went down. The FAA describes the El Dorado job as part of a years-long transition away from copper lines that are being disconnected as obsolete.
Daytime Work at Remote Sites Will Be Reevaluated
The router replacement happened in the middle of a Friday afternoon. The FAA says maintenance at major air traffic facilities normally happens at night, while replacements at remote facilities have been done during the day for decades without affecting its systems. “There will be reevaluation of when we do this work,” Duffy said. The FAA’s page repeats that it will reconsider when it schedules replacement and maintenance work. It also says it has no indication of a cyberattack, and that its new SMART tool, a data aggregator, does not feed the air traffic control system. The FAA says operators cancelled about 12,462 flights on 13 November, including 9,082 by major airlines, though not all were caused by the ground stop.
Why This Matters for Engineering Leadership
The change behind this outage was a single router at a single low-traffic site, and its effect reached 17 of 22 en route centers within hours. The scheduling rule treated the job as low risk because the site was remote, yet the risk came from what the device could send onto a shared network. Change risk depends on what a change can reach. For your own infrastructure, classify changes by blast radius on shared fabric, and treat any new device that can broadcast or open connections across that fabric as a change to every system on it. Ship new network hardware with chatty functions such as broadcast switched off until they are explicitly needed. Put connection-rate limits at the edge of each site so one location cannot flood the core. Keep a tested way to cut off a single site quickly, which is what stabilized the FAA network here. When a contractor installs equipment on your network, write the configuration baseline and the change window into the contract, and make sure your fleet inventory can find every similar device in hours, as the FAA did with its 60 routers.
FAQ
What caused the FAA nationwide ground stop on 13 November 2026?
The FAA says its legacy equipment became unstable when data flooded its FTI telecommunications network at 1,000 connection attempts every 4 to 5 seconds. The root cause investigation with L3Harris is still open.
Where did the data flood come from?
The FAA’s initial investigation points to two locations: a radar site in Amarillo, Texas, and a radio communications site in El Dorado, Arkansas, where L3Harris had been deploying a new router.
How long did the FAA ground stop last?
The FAA says the problem began at about 3:30 p.m. ET on 13 November and air traffic resumed at 8:15 p.m. the same day.
Was the FAA outage a cyberattack or caused by AI?
The FAA says it has no indication of a cyberattack or outside actor. It also says its SMART data aggregation tool is unrelated and does not feed the air traffic control system.
How many flights were cancelled?
The FAA says operators cancelled about 12,462 flights on 13 November, including 9,082 by major airlines, and notes that not all of them were caused by the ground stop.
What has the FAA changed since the ground stop?
It disabled the broadcast function on the originating routers and on about 60 similar routers, and it says it will reevaluate when it schedules replacement and maintenance work.
