Live data from Hacker News

Update on IT Security Incident at UCSF

ucsf.edu

11–20 of 150 posts

Re: Update on IT Security Incident at UCSF

#11
Wonder if the cost of backups for the affected servers would have been less than the US$1.1M they paid? ;)

At a guess, they could probably have put in a some fairly high quality storage + backups for that price, and still had money left over for all expenses paid staff vacations. :)

Re: Update on IT Security Incident at UCSF

#12
post #3

The paid ransom, will unfortunately embolden the criminals to strike again in search of the next big payday. If it worked once, it could work again.

Optimistic counterpoint: a high-profile, (relatively) high-value ransom payout like USC's may incentivise other orgs vulnerable to this kind of attack to take steps to prevent this kind of issue. Anything from restricting program capabilities/permissions for external executables, to keeping "colder" backups of business-critical data, to monitoring and responding to software that looks like it's traversing the whole f…

You can't really protect against this sort of thing. A lot of our IT security runs on trust. The only way to really prevent this is to make sure that ransom attacks don't pay out.

EDIT: I should mention that I've managed IT services for a major private university earlier in my career, and I am now a software security consultant. When I say it is not possible, I mean that pragmatically. A FAANG company can control their IT well enough to make sure this doesn't happen to them, but a hospital or university relies on computer systems running software way outside of their control. That MRI machine? Its controller is probably running some ancient version of Windows Server 2003 with proprietary drivers. That university registration app? Custom coded by generations of CS student interns running on a shared system whose operating constraints are set by the Novell GroupWise instance that is co-hosted on it.

As a practical matter, one of these organizations simply cannot reduce their risk to zero or near zero. There's too many attack vectors they don't have control over. The IT departments can't mandate proper security because they don't have the budget to enforce.

Re: Update on IT Security Incident at UCSF

#13
As a reminder, in 2017 UCSF offshored all of its IT staff to HCL Technologies and forced their then-employees to train their replacements before laying them off.

They brought the replacements into the Bay Area on H1B temporarily while they were trained by their soon-to-be-laid-off counterparts and then sent back overseas to continue their roles once training was complete.

https://sanfrancisco.cbslocal.com/2017/02/28/ucsf-tech-worke...

Re: Update on IT Security Incident at UCSF

#17

Earlier quoted context omitted.

Optimistic counterpoint: a high-profile, (relatively) high-value ransom payout like USC's may incentivise other orgs vulnerable to this kind of attack to take steps to prevent this kind of issue. Anything from restricting program capabilities/permissions for external executables, to keeping "colder" backups of business-critical data, to monitoring and responding to software that looks like it's traversing the whole f…

You can't really protect against this sort of thing. A lot of our IT security runs on trust. The only way to really prevent this is to make sure that ransom attacks don't pay out. EDIT: I should mention that I've managed IT services for a major private university earlier in my career, and I am now a software security consultant. When I say it is not possible, I mean that pragmatically. A FAANG company can control the…

You can't fully protect, sure, but you can have a person or team alerted on suspicious behaviour. You could also try to configure your infrastructure so that any code imported via vectors that ransomware usually uses is compartmentalized in a VM, container, or other chroot-like env.

And honestly, having even week-old cold backups makes this kind of attack _considerably_ less scary and cheaper, and it enables you to skip the payout (and I'm on the same page as you on that — if there's no money to be made, ransomware attacks will drop off).

Re: Update on IT Security Incident at UCSF

#18
What I find crazy about this -- no guarantee that the ransom payment would unlock the machines -- did they send 1.14M in one go or was it a smaller amount for the first machine, then an additional fee for each additional machine?

Also would be interested to know -- was it Bitcoin or some other cryptocurrency that was used?

Re: Update on IT Security Incident at UCSF

#19
post #6

Don’t they have insurance for these things? A small college near me had an attack like this but paid via insurance.

Would you want the insurance policy to pay out though?

At some level of recklessness, insurance becomes void. I think a lack of infrastructure to restore a hacked server — with data valued at over $1M — is negligent enough to not be covered.

But maybe UCSF are on MegaCo’s YOLO tier of server insurance, which is so expensive and isolated it has no impact on my MegaCo pet insurance premiums?

Re: Update on IT Security Incident at UCSF

#20
Maybe there should be a law that if you pay a ransom, you are required to pay the same amount as a fine. Because paying these ransoms is funding the criminals.... how about you have to also fund law enforcement to combat those criminals?

(also, this should reduce the amount that actually goes to the bad guys, since the amount of ransom would have greater downward pressure, i.e. if they'd probably not be able to collect more than $0.57M because that would cost UCSF $1.14M)

Post reply on HN