But if they did, people might die for real, if suddenly important information about patients is gone or manipulated. Or a important operation had to be canceled because of it. How can one critizise bombings of hospitals and then potentially do the same on a software level?
Get over yourself.
I think the main concern of the OP was with _human_ collateral damage that can result from the deleted bits. Having worked for years in EMRs, I can definitely say that sustained IT outages in modern hospital systems can and often do lead to increased deaths.
I usually oppose “whataboutism” but I can’t bring myself to do it here. This is a war where the nation of Russia has decided to murder Ukrainians. Means to interfere with Russia’s blood-soaked campaign must be viewed not in isolation, but in the context of what they are attempting to stop.
To be clear, I am not making a general purpose “the end justifies the means” argument. I am making a narrow and specific “the circumstances justify extreme behavior with severe externalities” argument.
Yes, this is a warcrime. But the proposed solution is to react with a warcrime?
I do not agree.
Now, there's nothing this action could achieve aside from making people angry, becoming Russian propaganda-fodder, reducing trust in NPM, etc, in any remotely likely scenario. So any damage is a crime, and arguably even the possibility of damage is a crime.
* A strike on head of ISIS while being surrounded with a class of children might pass this test. Similarly even if one side (say ISIS supporters) has more civilians killed that does not imply a crime by itself.