They pointed out their MRI machine depended on an Windows Server 2K3 machine, which they didn't patch, networked on the corporate network (to access/transmit results). I asked why they couldn't patch or upgrade it, to which they responded:
> "Every time we change patch or change the software configuration we need to have it re-certified before we can use it on a patient, and that takes months and $10,000's"
I really hope we'll see some movement in the health security space, maybe leveraging SDN/Micro-segmentation or Identity-Aware-Proxy software like Hashicorp Boundary, but these aren't (really) zero-cost and add friction/delays to accessing the information you need about your patient... Time will tell.
And I can't really agree with everyone solely blaming the attackers. Bad people exist. You don't leave your front door wide open when you leave your house either and then act all surprised if you get robbed. I don't really let "don't blame the victim" count here. A malware attack is not like a natural disaster that just happens. Don't build your house on quicksand.
Maybe it's actually good this happens now and that it's "just" ransomware. Because maybe it raises awareness and leads to better security. Next time it might be a coordinated attack by a state actor that doesn't just have money on their mind.
They're still treat like Demi G-ds and can and do demand password-less access since they can not be bothered with anything that is not billable, err, in the patient's best interest.
The amount of unpatched crappola, the self bought software, of alternative inbound connectivity and hugely expensive machines from well known vendors without even the slightest form of IT security is mind boggling.
Add to that the horrendous state of "digital health" which is in its core based on standards like HL7, makes me 100& sure that what we are seeing now it just the humble beginnings.