Over five days in November, Microsoft services went down, flights stalled across the US, and JPMorgan Chase hit a payment snag. People on X immediately blamed an "AI hack," yet a Quartz breakdown shows the facts are much simpler.
There's no trace of hackers, and the three incidents aren't linked. Microsoft was tripped up by a network update. Flights stopped because of a new router in southern Arkansas from contractor L3Harris Technologies, according to Transportation Secretary Sean Duffy, with the root cause still under review. At Chase on November 16, a bug slowed some transactions and affected payments through Bank of America, Citigroup, and Zelle, though account access stayed up and the issue was fixed the same day. On top of that, a public bank must disclose cyber incidents to the SEC or face penalties.
So where does AI come in? YouTuber Nate Herk points to an indirect effect: in 2026, Anthropic gave banks its Mythos model, which Herk says uncovered over 10,000 flaws. Bank of America CEO Brian Moynihan confirmed the scale of the scramble to CNBC, as reported by Zafran: the bank hired 1,000 contractors alongside 3,000 staff engineers to patch software after checks like Mythos.
Herk notes there's no confirmed link between patching and the Chase outage. Still, the logic is easy to follow: banks run decades-old legacy code full of fragile links, and rushing to patch holes on the fly makes it easy to trigger a new failure.
