Top
Best
New

Posted by zulln 5 hours ago

FAA issues national ground stop(www.nbcnews.com)
https://www.fly.faa.gov/adv/adv_otherdis?adv_date=11132026&a...

https://www.nytimes.com/live/2026/11/13/us/ground-stop-natio...

1348 points | 832 commentspage 19
erickhill 4 hours ago|
Took themselves off floppy disks and now look what happens. /s
consumer451 4 hours ago||
[flagged]
rooftopzen 4 hours ago||
The government's AI tool company likely had no understanding of context window limitations (although could be Iran or a false flag to be blamed on Iran) https://www.flyingmag.com/faa-smart-ai-now-active-at-22-artc...

Company that created the tool (called Airspace Intelligence) is interesting in terms of demographics for these types of companies - implies lots if the tool 'does not work as designed' (which it will obviously not)

https://www.linkedin.com/company/airspace-intelligence/peopl...

jbredeche 5 hours ago||
Maybe related to this? https://gizmodo.com/the-faa-is-rolling-out-its-ai-air-traffi...
aschla 4 hours ago||
Doubtful. That tool likely suggests "improved" clearances, which a controller can use at their discretion. It likely does not control clearances directly.

And the early reports describe the issue as related to ATC communication infrastructure.

eli 5 hours ago|||
No. A route planning tool, AI or not, couldn't possibly cause this.
sigmarule 5 hours ago||
Oh man. It would be such a perfect encapsulation of these times. If not related, then truly cursed launch/PR timing.

It looks like the FAA started testing it in DC in September and annouced a nationwide rollout three days ago (https://www.faa.gov/newsroom/statements/general-statements). This makes it seem more unlikely to be the cause, though:

  A team of FAA specialists in the SMART center in Washington, D.C., analyzes and validates the SMART data and provides scheduling recommendations to the agency’s Air Traffic Control System Command Center in Warrenton, VA. The Command Center then routes these recommendations to air traffic control facilities for local leadership to consider.
Of course, these may be the ATC analog of claude code approvals, or they may be overstating the level of human oversight/HIL - but the overall situation seems likely more serious/unclear than something like this. I'm not sure I'd call vibe air traffic control going haywire an "extraordinary" circumstance, though certainly extraordinarily stupid.
etaioinshrdlu 5 hours ago||
Can't help but suspect that "AI" is to blame and the response is to sort of ban open source AI, or owning too much compute!
gdulli 5 hours ago|
No, and when Netflix goes down it's not a false flag due to their fear of people running their own Plex servers. Perhaps you just have anxiety about losing something you depend on, so you're projecting.
LPisGood 4 hours ago||
Netflix isn’t widely reported to be hacking systems left right and center.
hector_vasquez 4 hours ago||
I wonder if it was Claude or ChatGPT? My money would be on the latter. Anthropic seems to generally have better security and guardrails, and apparently OpenAI’s newest cyber model is extremely capable.
ChrisArchitect 5 hours ago||
No longer exists? Temporary glitch? (Edit: OP url was wrong https://www.fly.faa.gov/ois/oisedit/adv_oisedit_dis?number=1...)
lazystar 5 hours ago|
its front page news everywhere right now. something definitely happened
delis-thumbs-7e 5 hours ago||

  ...And Claude, remember NOT to make NO mistakes.
Let's just use GPU's to play Dark Souls like god intended and stop this nonsense.
deadbabe 5 hours ago||
This is it, massive cyberattack on the USA.
soundworlds 5 hours ago|
I've worked in IT long enough to know it is 99/100 times an internal glitch or misconfiguration, not hackers, that brings a system down.
drillsteps5 4 hours ago|||
Two major completely unrelated outages, 2 days in a row?
Aboutplants 5 hours ago||||
“Jimmy unplugged the wrong cable”
hermitdev 4 hours ago||||
This week alone, I've seen multiple servers fail due to hardware issues: multiple bad fans, a failed hard drive, bad network optics. Another server got bumped while an adjacent server was being worked on and lost time sync, leading to software instability/crashes. We don't have a large fleet, but we do have more servers than employees, and issues add up. We don't need any help with instability, entropy has us covered, thank you very much.
jgerrish 4 hours ago||||
The nifty thing about stories is that they don't have to be about something directly to trigger strong feelings in your crowd or listeners.

When we are dealing with the FCC, grounding all flights and radio failures in open bands, you can drive a lot of changes.

There are a lot of pesky ESP32s out there, right? Did your company choose to go with the Espressif platform? You know, the hackable SDR one that crashes planes and clubs baby seals and tries to sell your grandparents bad financial scams?

deadbabe 4 hours ago|||
Ok, and the Microsoft outage yesterday?
cyanydeez 4 hours ago|
Yes, of course, who wouldn't just believe random statements from the worst prepared admin in history of America.
More comments...