I'm unconvinced this is true.

Most of what libraries do technologically is now outsourced to SAAS companies that specialize in the library space (there are at least 2 major one in the US, and I think at least 3 or 4). (encore and bibliocommons are the 2 major ones that I remembered and am pretty sure I've seen at least 2 others).

If everything you are doing is in the cloud, I'd argue that recovering from a ransomware attack is more of a question of if the SAAS company did its job right (which it should be more capable of handling). Of course, that's a reasonable question if they did (similarly why I prefer paying with paypal or google or even (as not so common) amazon third party payment systems than a retailers own internal payment system, yeah they might screw up too, but they invest a lot more into not screwing it up).

Well, i am basing this on 25+ years of being sysadmin at governmental organisations where most of my colleagues never learned the password to the next level. The password being “no”. Instead they always (alway!) hide the pain so the management layers above know nothing. If they were ever interested. By showing the pain we may actually get them interested. We tried hiding it. Didnt work. Lets try something different.

Closing things down is the argument taken to a logical extreme. Could they function without collecting that much PII though? If you can't do it right, don't do it, is far more reasonable when limited to just the collection of PII or other potentially sensitive information. Some identifying data might be absolutely required of course, but the great many cases like this show that collecting the absolute minimum must become the default choice.

If you leave your car unlocked and police detects that, they can confiscate your keys and probably even fine you after you collected them. The same principle applies to everything. To a certain extend, it is your duty to secure things. Just putting a sign there "please dont fuck with me" isn't enough these days.

The relevant part isn’t whether they can guarantee security, it’s whether they followed best practices in attempting to ensure it.

No restaurant can “guarantee” that you won’t die from eating their food, but that doesn’t mean anything goes in the kitchen.

Not really. The victim here is the people whose information has been distributed, not the system (and the people running it) the information was exfiltrated from.

If we are going to use an SA analogy, this is closer to the recent case of a hotel chain where there were several instances of staff giving “randoms” keys to rooms that they should not have. Best practise procedures were not being followed, the ne'er-do-wells took advantage of this, someone else suffered as a result.

If you process personal data, you must secure it against attacks. Are you suggesting that they could just put the addresses of their customers on their website with a warning not to download them, and if someone downloaded them anyway then it would be the fault of that person that they were leaked?

> Somehow its only computers where people turn off their brains and argue that proactive measures need to be 100% effective at blocking all attacks no matter how high the cost.

I don't think I've ever seen anyone say they should be 100% effective. Even those physical locations follow some physical security practices. The difference is that computers can be attacked from literally anywhere in the world, and good luck finding the people who did it.

> Or alternatively we could hold criminals responsible

Ok. Where are the criminals who did this?