Neglect, then panic!
https://www.youtube.com/watch?v=4-QnJLTblY8&list=UU9rJrMVgcXTfa8xuMnbhAEA - video
https://pivottoai.libsyn.com/20261118-this-weeks-us-bank-outage-was-ai-related-after-all - podcast
time: 5 min 41 sec
Neglect, then panic!
https://www.youtube.com/watch?v=4-QnJLTblY8&list=UU9rJrMVgcXTfa8xuMnbhAEA - video
https://pivottoai.libsyn.com/20261118-this-weeks-us-bank-outage-was-ai-related-after-all - podcast
time: 5 min 41 sec
This is also going be the same issue which has iirc happened before. People who write automatic bug finding software (like ‘network intrusion detection’ software, which tended to create a lot of useless reports) are at this immature stage incentivized to create the max amount of potential bugs. As that is how people will evaluate the tools. One bots finds 400 bugs 80% cve of 9+ vs other tool that finds 40 none 9+. With nobody around to evaluate 440 bugs, the higher number is better.
(Note im speculating here based on how it went in the past).
Reminds me of twenty years ago when the old folks would download Super Malware Scanner from a banner ad and it would find 241 SEVERE viruses which it would “quarantine” if only you would buy a subscription.