BREAKING · UPDATED. Every departing flight in America was grounded for five hours on Friday. Officials say it wasn't a cyberattack — it was the modernization work itself.
IT WASN'T A CYBERATTACK. IT WAS AN UPGRADE. On Friday, November 13, 2026, at 20:54 UTC — 3:54 p.m. Eastern — the FAA issued a nationwide ground stop: every departing flight in the United States ordered to stay on the ground.
It lasted roughly five hours. More than 10,000 flights were delayed. Thousands were canceled the next day. The FAA said it expected full recovery by Sunday.
And the cause, according to Transportation Secretary Sean Duffy, was the agency's own modernization work — new technology overwhelming old infrastructure it was supposed to replace.
WHAT BROKE, SPECIFICALLY. Duffy's account, given Saturday, is that the trouble began with a contractor replacing copper wire with fiber at one facility, as part of updates to an older communications system managed by L3Harris Technologies.
The new equipment didn't fail. It overwhelmed what was already there — outdated routers and switches that couldn't handle the load. Homeland security officials described the same thing: new technology pressuring aging infrastructure until it gave.
The detail that makes it a national event: the failure hit radios and radars used by high-altitude air traffic controllers — the Air Route Traffic Control Centers, the regional facilities that handle traffic above the airports, not the towers. If those centers can't talk to aircraft, nothing new can take off anywhere.
WHAT IT WAS NOT. Officials were explicit on two points, and both matter:
Not a cyberattack — no indication of one.
Not connected to the Microsoft Outlook outage earlier in the week. Two big technology failures within days of each other, and the government's position is that they're unrelated.
Which leaves the least dramatic explanation, and the most embarrassing one: the system was being modernized, and the modernization broke it.
HOW OLD IS OLD. The FAA has been trying to replace its communications systems for years; some of that equipment is 20 to 30 years old. A radio network built for a smaller, slower aviation system doesn't fail because someone types the wrong thing — it fails because it's being asked to do something it was never built to do.
And that is exactly the shape of the failure here: the new thing worked, and the old thing couldn't carry it.
THE MONEY, WHICH IS THE REAL STORY. Within a day, Duffy had put a number on the overhaul: $50 billion — up from the $30 billion the administration had been asking Congress for.
Read the sequence. The modernization program is what caused the outage. The outage is now the argument for funding more of the modernization program — at $20 billion more than the number on the table a day earlier.
That's not hypocrisy. It's how infrastructure funding actually works: the money goes where the pain is visible, and nothing is more visible than every airport in the country stopping at once.
But it does mean the one thing nobody has to answer for is the failure itself — because the failure is what unlocks the money.
THE RECOVERY IS THE HARD PART, NOT THE STOP. A ground stop is a decision; a recovery is physics.
Planes and crews were in the wrong cities. A cancelled departure doesn't just cancel a flight — it strands a crew and a tail in the wrong place, and every one of those has to be repositioned before the schedule can run again. That's why the cancellations kept coming on Saturday, and why the agency was talking about Sunday.
Tens of thousands of people were stranded. The stop lasted five hours. The disruption lasted days.
IF YOU WERE ON ONE OF THOSE FLIGHTS: THE REFUND IS AUTOMATIC. Under the Department of Transportation's rule finalized in April 2024, airlines must issue prompt automatic cash refunds when they cancel a flight or make a significant schedule change and the passenger chooses not to travel.
Three things people get wrong: you do not have to ask for it; you do not have to accept a voucher or a credit; and the refund goes back to the original form of payment. A cancellation caused by an air traffic control failure is still a cancellation.
What isn't automatic is everything else — hotels, meals, the rental car you didn't need. Those depend on the airline's own policy, not federal rule, which is why two passengers on the same cancelled flight can be treated completely differently.
SPLASH'S READ: The first nationwide ground stop since September 11, 2001 wasn't caused by an attack or a storm. It was caused by routine maintenance on a system that was already too old to take it.
That's the part worth keeping: a country builds critical infrastructure, runs it for thirty years, and then discovers that the act of replacing it is itself the risk — because you can't take the system offline to fix it. You have to upgrade it while it's flying.
Nobody did anything reckless here. A contractor pulled copper and ran fiber. The FAA applied patches. The old machines couldn't hold the new ones.
And for five hours, the busiest airspace on earth stood still.