[deletes]

It would spend all it's time and resources in order to even slightly decrease the probability of a human coming to harm.

You are assuming there are no thresholds, which is not correct for any decent ( fictitious )ai, I believe.