AT&T 1990: The Software Bug That Spread Like Fire #shorts

Added: 2026-05-17

[00:00:00]Imagine making one phone call and accidentally triggering a chain reaction across America's phone network. That's what happened to AT&T in 1990.

[00:00:09]Back then, long-distance calls depended on massive electronic switching systems.

[00:00:14]These machines decided where calls should go, how they should connect, and how the network should recover if something went wrong.

[00:00:22]But hidden inside a software update was a bug.

[00:00:26]And when one switch crashed and rebooted, it sent a recovery message to nearby switches.

[00:00:32]That message was supposed to help the network stabilize, but instead [music] it triggered the bug.

[00:00:39]The nearby switches crashed, too.

[00:00:42]Then they rebooted.

[00:00:44]Then they sent the same message to other switches.

[00:00:47]And the failure started spreading.

[00:00:49]One switch, then another, then dozens more.

[00:00:54]Soon, parts of AT&T's long-distance network were stuck in a loop, crashing, rebooting, and triggering even more failures.

[00:01:02]And the scary part is this was not caused by hackers.

[00:01:06]It wasn't a physical attack. It was a software mistake inside the system that millions of people depended on.

[00:01:14]For about 9 hours, long-distance calls across the country were disrupted. More than 50 million calls were blocked.

[00:01:21]Engineers eventually stabilized the problem, but the damage was already done.

[00:01:26]A bug hidden inside a recovery system had turned the network's own safety behavior into the [music] threat.

[00:01:33]And that's the terrifying lesson.

[00:01:35]Sometimes technology fails because it crashes.

[00:01:39]But sometimes it fails because it tries to recover the wrong way.

[00:01:44]So, here's the question. If one recovery message could break a phone network, what happens when today's systems are even more connected?