Scalable systems need to use heuristics to catch threats. Needing concrete evidence in every case means that an enormously higher amount of malicious resources will not be flagged.
There is a policy argument as to the right balance of concerns here. But there is a clear trade-off to make.
Giving everyone a fair trial just doesn't scale. It costs too much.
"Your Honor, we banned this person's website because his web page contained the word 'bitcoin' more than 5 times" will not hold up.
"Your Honor, we banned this person's website because it contains a bitcoin miner script. See, here is the script, and it matches the hash value found in these other attacks" hopefully holds up.
In other words, if you can't deal with the false positives in a timely manner. You SHOULD be liable for the damages.
I can't build a budget car put together in an unsafe manner. Then complain I can't compete due to all the peoples cars crashing and blowing up and suing me.
Libel suits can be financially catastrophic, so even a tiny false positive rate could present risk that disincentivizes producing such software at all.
And a threat detection mechanism that has a 0.0% false positive rate is conservative to the point of being nearly useless.