I've seen a few people say it but i don't know how likely it is. Microsoft, airlines etc in days is a bit weird

you are viewing a single comment's thread
view the rest of the comments
[–] 2 points 1 day ago (2 children)

I think the FAA ground stop is the most likely to be a cyber attack.

They’re trying to cover their ass by spitting out generic terms like “communication disruption” and “infrastructure failure”. Which IMO is what you say when you don’t want to create panic or piss people off. So that means one of a few things:

  • The AI tool they bragged about implementing fucked everything up.
  • People that got let go as a result of DOGE weren’t there to fix the problem.
  • They did some other sort of colossal fuck up that makes those in charge look bad. (If it was just some dude who cut a cable they would have thrown him under the bus)
  • There was a genuine cyber attack.

I wasn’t leaning toward the last one until the FBI apparently got involved in the investigation.

As for the rest:

Microsoft admitted they fucked something up internally and how it cause the outage.

Banking sites was probably a run of the mill outage.

NSA outage today (if real) was probably an internal screw up too.

  • source
  • hideshow 2 child comments
  • [–] 4 points 1 day ago

    The problem is when you speak to people who've worked on the FAA networks, they verify that some of the systems are old and predate Ethernet. Is that to say they can't be the target of cyber attacks? No. But Occam's Razor applies...

    But I've worked on similar systems and drifting clock/timing for said equipment would definitely cause transmissions arrive out of sync and corrupted, easy. The cascading system collapse was what drove us to abandon the old equipment, as we had over a dozen sites that had to be powered down to recover from the issue or it would just propagate all over again.

    To me, the explanation doesn't seem too outlandish. They keep saying it's the TDM systems, but everyone is blaming AI. I don't think most people are knowledgeable enough about networks outside of IP, so the assumption is it's switches, routers and computers that everything else uses. That's not the case, and I have my doubts.

    My concern is mostly with redundancy: they should have multiple backup systems to prevent this from occurring. The fact that all of them failed is a bigger issue. I think this is a result of brain drain and antiquated systems rather than systemic incompetent AI tech bros. This doesn't mean both aren't happening (and arguably even if it's the former, it's still driven by the latter) - but i would pause on the AI judgement briefly because that doesn't account for the real issue at hand.

  • source
  • parent